Separating abusive from efficient related-party transactions: evidence from India

Abstract

Purpose

The aim of this study is to understand a family firm's choice of related-party transaction (RPT) types and analyze their value impacts to separate the abusive from benign RPTs.

Design/methodology/approach

It uses a 10-year panel of BSE-listed 378 family (and 200 non-family) firms. The fixed effects, logit and difference-in-difference (DID) models help examine value effects, propensity and persistence of harmful RPTs.

Findings

Loans/guarantees (irrespective of counterparties) destroy firm value. Capital asset RPTs decrease the firm value but enhance value when undertaken with holding parties. Operating RPTs increase firm value and profitability. They improve asset utilization and reduce discretionary expenses (especially when made with controlled entities). Family firms have larger loans/guarantees and capital asset volumes but have smaller operating RPTs than non-family firms. They are less likely to undertake loans/guarantees (and even operating RPTs) and more capital RPTs vis-à-vis non-family firms. Family firms persist with dubious loans/guarantees but hold back beneficial operating RPTs, despite RPTs being in investor cross-hairs amid the Satyam scam.

Research limitations/implications

Rent extractability and counterparty incentives supplement each other. (1) The higher extractability of related-party loans and guarantees (RPLGs) dominates the lower extraction incentives of controlled parties. (2) Holding parties' bringing assets, providing a growth engine and adding value dominate their higher extraction incentives (3) The big gains to the operational efficiency come from operating RPTs with controlled parties, generally operating companies in the family house. (4) Dubious RPTs seem more integral to family firms' choices than non-family firms. (5) Counterparty incentives behind the divergent use of RPTs deserve more research attention. Future studies can give more attention to how family characteristics affect divergent motives behind RPTs.

Practical implications

First, the study does not single out family firms for dubious use of all RPTs. Second, investors, auditors or creditors must pay close attention to RPLGs as a special expropriation mechanism. Third, operating RPTs (and capital RPTs with holding parties) benefit family firms. However, solid procedural safeguards are necessary. Overall, results may help clarify the dilemma Indian regulators face in balancing the abusive and business sides of RPTs.

Originality/value

The study fills the gap by arguing why some RPTs may be dubious or benign and then shows how RPTs' misuse depends on counterparty types. It shows operating RPTs enhance operating efficiencies on several dimensions and that benefits may vary with counterparty types. It also presents the first evidence that family firms favor dubious RPTs more and efficient RPTs less than non-family firms.

Citation

Saurabh, K. (2023), "Separating abusive from efficient related-party transactions: evidence from India", Asian Review of Accounting, Vol. 31 No. 4, pp. 631-657. https://doi.org/10.1108/ARA-06-2022-0136

1. Introduction

According to agency theory, the conflict-of-interest hypothesis views related party transactions (RPTs) as the tombstone of shady governance practices to expropriate value (OECD, 2014). Hidden motives and verification problems of RPTs allow firms to mislead with impunity in markets like India, characterized by the prevalence of family firms and poor investor protection. Recent corporate scams in India [1] have brought RPTs into the cross-hairs, inviting calls for a ban on RPTs. Indian Ministry of Finance (2012) regards related party loans or buy/sale of assets as the favored method of expropriation. On the other hand, the efficient contract perspective views RPTs as efficient contracts to reduce transaction costs, fill the institutional gap and impart scale/scope economies (Ryngaert and Thomas, 2012). A clampdown can prevent efficiency gains from RPTs. This two-sidedness of RPTs has led to policy dilemmas. As such, identifying value-reducing RPTs and related parties' incentives (e.g. loans to directors) may increase the efficiency of RPT laws.

Several studies examine the value relevance of RPT items, mainly on the cash flow direction, but find ambivalent results (Cheung et al., 2006; Peng et al., 2011; Wong et al., 2015). The mixed evidence may owe to four reasons overlooked in past studies. First, studies have selectively used RPTs such as other receivables (Jiang et al., 2010), guarantees (Berkman et al., 2009), asset purchases (Cheung et al., 2006) and asset sales (Wong et al., 2015) suited to the local context. Except for a few studies (Bona-Sánchez et al., 2017), very few have thoroughly studied the value effects of composite RPT types and explained why they might differ in value effects. Second is the lack of attention to tunneling incentives of counterparties, especially since parties controlling the firm have higher tunneling incentives than parties under the firm's control (Claessens et al., 2000; Johnson et al., 2000). Third, while family firms are considered more likely to use RPTs, RPTs by family firms remained underexplored in the prior literature. Hence, research has not provided evidence that family firms may favor operating RPTs to build competitive advantages. Fourth, most articles study Chinese firms where the nature of the political organization drives tunneling/propping (Cheung et al., 2009a, b; Jian and Wong, 2010). Applying this specific evidence to RPTs in other markets is challenging. I address these concerns and analyze the value impacts of RPT types and counterparties to help separate the abusive from benign RPTs.

The value relevance of RPT types may differ on value extractability (recurrence, detection evasion and extraction size). The operating RPTs (buy/sale of goods and services) are related to routine operating income/expenses that are hard to manipulate, get settled the same year and are easy to verify. Whereas non-operating RPTs like related-party loans/guarantees (RPLGs) and capital asset RPTs, are infrequent, remain untied to current operations and can be concealed. Since some RPTs can be efficient (Ryngaert and Thomas, 2012), I ask whether non-operating RPTs by family firms destroy firm value and whether operating-RPTs are related to increased firm value.

The value extraction also hinges on tunneling incentives of counterparty types. The ownership and control divergence literature suggests that holding parties (owners/directors/relatives/holding companies) carry greater tunneling incentives than those below in the control chain of a group (Claessens et al., 2000). While controlling owners can pocket the entire loan from a listed related company, operating RPTs with controlled parties (associates/joint ventures/subsidiaries) offer much smaller expropriation size and higher detection chances. Hence, I explore whether some RPT types with holding parties are more likely to destroy firm value than RPTs with controlled parties and whether operating RPTs with controlled parties increase firm value. Further, I question whether operating RPTs will help improve a firm's operating efficiency.

A top-five nation by population, GDP and market cap, studies of corporate governance of Indian firms have largely not appeared in research. Indian data bring novel views on all RPT types with counterparties – data unavailable in many markets - that may provide further new evidence and help literature grow. A sizeable number of non-family firms (200 in my sample) helps differentiate the propensity of family firms to undertake RPTs as conjectured (without evidence) in the literature. Since India inherits common law and institutions from the UK yet retains its local idiosyncrasy, it is an excellent setting to test the transplant of corporate governance into emerging markets (Afsharipour, 2009). As such, it offers contrasting impulses for productive uses of operating RPTs and appropriate value through non-operating RPTs, unexplored previously.

Three main reasons facilitate expropriation through RPTs in India. The first is the prevalence of family firms, concentrated ownership and groups/cross-holdings. The second is the strength of family logic. In the West, family firms behave like non-family firms, with family owners increasingly taking a more passive role as investors. But a family tends to dominate the firm in India by holding CEO/board positions. The third is weak law enforcement. RPTs are common, happen in concealed terms, boards are pliant and laws are feeble in critical areas like group-level disclosure, arm's length criteria and the majority of the minority rule (OECD, 2014).

However, RPTs have been valuable in bridging institutional voids and raising funds for growth for decades (Khanna and Palepu, 2000). Also, since Indian family firms strive to preserve their legacy and reputation (Piramal, 2010), they may appropriate in small amounts only (Kang, 2015). A high ownership concentration (50% average) reduces their incentives to steal (Durnev and Kim, 2005). Standout investor protection laws (World Bank, 2008), independent media and democracy in a common-law nation amplify investor voices. Unlike East Asia, where collectivism promotes opportunism (Chen et al., 2002), India retains robust individualism (Sinha et al., 2001). Institutionally and culturally, it sits between the West and the East, furthering generalizability.

I use a panel of 378 largest family (and 200 non-family) firms ranked on market cap. I mainly analyze the family firm sample and rely on fixed effects models to control firm heterogeneity. While controlling for year and industry effects on firm value (as Tobin's q), I also employ lagged RPT levels to account for lagged effects of loan/guarantees and capital asset RPTs. I also use a difference-in-difference (DID) model around the Satyam scam in 2009, culminating in external director exits from several firms, to analyze family firms' behavior toward RPTs.

First, I show that RPLGs and capital RPTs reduce the firm value, but operating RPTs enhance value. Second, I study how counterparty incentives affect extraction. RPLGs destroy value irrespective of counterparty types. However, capital RPTs with holding parties are valuable, and operating RPTs with controlled parties increase asset utilization and reduce sales and operating expenses, increasing profitability and value. Third, I study whether family firms prefer making opportunistic RPTs. I show they undertake higher volumes of opportunistic RPLGs and capital RPTs but smaller volumes of operating RPTs than non-family firms. However, they are less likely to undertake RPLGs (and even operating RPTs) and are more likely to use capital RPTs than non-family firms. Also, the DID analysis shows that family firms prefer to persist with abusive RPLGs but reduce visible operating RPTs, even when the market closely watches RPTs by firms.

This study contributes to the literature on RPTs in three ways. First, while research tends to concur about opportunism behind RPLGs, very few studies examine capital assets or operating RPTs (Bona-Sánchez et al., 2017; Kang et al., 2014). Others (Cheung et al., 2006, 2009a, b; Wang et al., 2019; Wong et al., 2015) have studied it, bifurcating these on cash flow directions, with mixed evidence. Instead, following a few studies (Bona-Sánchez et al., 2017; Kang et al., 2014), I assume that both the sell and buy sides are susceptible to extraction. I confirm the efficiency perspective by finding novel evidence on how operating RPTs by family firms increase asset utilization and reduce operating expenses and SGAs, augmenting firm profitability. I also show fresh evidence of asset RPT's positive value effects. This way, I move research beyond the already known positive or negative value effects (largely mixed across different markets and institutions) of RPTs.

Second, it is among the first few studies to examine counterparty incentives and substantially qualifies tone-at-the-top (Kohlbeck and Mayhew, 2017) and the control-ownership divergence literature (Claessens et al., 2000). It yields some novel findings that extractability and counterparty interactions supplement each other. (1) The higher extractability of RPLGs dominates the lower extraction incentives of controlled parties. (2) Holding parties' bringing assets, providing a growth engine and adding value dominate their higher extraction incentives. (3) The big gains to the operational efficiency of operating RPTs come from RPTs with controlled parties, generally operating companies in the family house. Unreported results also show that controlled parties majorly contribute to reducing operating expenses and holding parties in reducing SGAs.

Finally, I find the first evidence beyond conjecture comparing family and non-family firms' RPT preferences. RPLGs and capital asset RPTs seem more integral to family firms' choices than non-family firms. Family firms undertake bigger RPLGs (more likely to cross the materiality), higher capital asset volumes and much smaller operating RPT volumes than non-family firms. However, they seem to prefer RPLGs for specific purposes as they are less likely to advance RPLGs in any year than non-family firms. Instead, they are more likely to undertake capital RPTs and less likely to use operating RPTs. Their choice of RPTs depends on their intended use, as inferred from the value effects of RPT types. They reduce valuable operating RPTs yet continue using dubious RPLGs, despite RPTs being in regulatory cross-hairs amid the Satyam scam.

My study has important practical and policy implications. Investors, auditors or creditors must pay close attention to RPLGs as a special mechanism to use shell companies to siphon funds. Auditors should intensively scrutinize capital RPTs as they remain a gray mechanism offering both benefits and extraction opportunities. RPLGs and capital RPTs require more regulatory attention. While operating-RPTs do benefit firms, curbing RPLGs alone may push exploiting operating RPTs to extract value. Reinforcing physical verification in audit works may deter phantom operating RPTs. In sum, the findings should interest regulators, investors, creditors and auditors.

Section 1 presents the introduction; section 2, the institutional background; section 3, the literature review and hypotheses; section 4 explains data and variables; section 5, results and discussion; and section 6 is the conclusion.

2. Institutional background of India and RPTs

India forms a vital setting to study the tension between convergence to the US/UK model and local idiosyncrasy (Afsharipour, 2009), revealing contrasting views behind RPTs.

Family firms are prevalent in India. About 65% of the top 600 firms are family-owned. Families hold dominant control, pack the board with relatives and form groups connected by pyramids on par with South Korean firms (Masulis et al., 2011). But its political, economic and social rigidity has ensured less evolved institutions block the convergence beyond formalism and characterizes high information asymmetry, high contracting costs, weak courts and poor state capacity to protect investors (Afsharipour, 2009). Not surprisingly, Bertrand et al. (2002) find that RPTs present severe conflict between owners and investors and encourage tunneling in India.

The Companies Act, 1956 (sections 297, 299, 300), Clause 49 and the Accounting Standard (AS18) regulate RPTs and require board approval when directors are interested. India quickly tailed the UK/US reforms closely. Clause 49 brought the RPT rules closer to the US/UK standards. Companies Act 2013 (section 188) brings a stricter RPTs regime. However, while laws requiring disclosure of materiality are in place, they offer little investor protection and effective means of redressal (OECD, 2014). The disclosure happens with lag, the audit committee reviews RPTs with lag, the majority of the minority rule is absent and group-level disclosures are not mandatory (OECD, 2014). These drawbacks exacerbate RPTs' abuse as asset siphoning becomes easy. Pyramids and cross-holdings compound the problem. The set of unlisted firms owned by each family with no disclosure requirements increases the opacity of RPTs to evade the law. As a result, despite strict regulation, RPTs comprise a large part of business activities in India.

But as a common law country, India draws heavily from the UK Companies Act, 1948, the Cadbury report and US SOX. It has built up its corporate governance and often receives a high ranking (at 33) in investor protection (World Bank, 2008). Indian firms voluntarily initiated governance reforms and have moved fast (Afsharipour, 2009). Competitive politics and media strengthen investors' voices. A good case was the Satyam scam when media and investor pressure forced the board to recant, and the promoter to accept fraud. Though courts deliver judgments with inevitable delays in India, they have been hard on scams offering no easy escape.

The frequency, structuring and extraction attractiveness of some RPTs may get shaped by institutional settings. Historically, emerging markets have lacked institutions to step up economic growth. Group structure has been extremely helpful in bridging institutional voids, raising funds and gaining scale/scope economies (Khanna and Palepu, 2000; Khanna and Yafeh, 2007), reducing risks and protecting from distress (Gopalan et al., 2007). For example, Chaebols expedited industry formalization and industrialization in South Korea. In India also, RPTs have facilitated industry expansion - vertically and horizontally - and created industrial growth over the years. Local firms seek to raise capital, get licenses and meet eligibility criteria as entrepreneurial opportunities emerge in the industrial void. Related firms can help aggregate capital/assets to help enter new businesses and provide trade relationships on lenient terms. While capital assets RPTs with owners/holding companies can be a prime mechanism for firms to raise/dispose of capital assets, operating RPTs with associates/subsidiaries can enhance the productive capacity of firms.

Family firms may also promote efficient RPTs as firm survival is crucial to family owners, who have their wealth and human capital concentrated in the firm. These owners accord high priority to sustaining the business/legacy (Bertrand and Schoar, 2006) and may avoid practices that might destroy their reputation (Deephouse and Jaskiewicz, 2013). Indian family firms also strive to preserve their legacy and reputation (Piramal, 2010) and may employ RPTs efficiently. In this light, it appears reasonable that operating RPTs by Indian family firms may be valuable.

3. Review of prior research

There are two opposite views in the literature about RPTs. The efficient transaction view suggests that RPTs boost operating efficiencies and increase firm value (Ryngaert and Thomas, 2012). In this view, RPTs reduce transaction costs, help overcome institutional void (Khanna and Palepu, 2000) and impart scale/scope economies (Fisman and Khanna, 2004). However, the conflict-of-interest view suggests that in markets like India that exhibit poor corporate governance, controlling owners (e.g. family owners) can use RPTs to tunnel value from the firm (Djankov et al., 2008; Johnson et al., 2000), making investors depreciate firm value.

3.1 The nature of transactions

Based on the nature of transactions, RPTs are mainly of three composite types: buy/sell of goods and services, buy/sell of capital assets and lend/borrow loans and guarantees. Cheung et al. (2006) further divide each composite RPT type based on cash inflow or outflow direction, where outflows indicate tunneling. The literature has mainly studied RPTs on a cash flow basis and finds mixed results. For example, regarding the buying/selling of goods and services, prior works show related party sales with firms in the same or related industry in Taiwan (Wang et al., 2019) or with group firms in China (Wong et al., 2015) improve firm value. Cheung et al. (2006) find that both buying and selling goods/services (separately) in Hong Kong show a null effect. However, Ge et al. (2010) show that related party sales reduce firm value in China. Regarding capital RPTs, Cheung et al. (2006) find that both asset purchases and asset sales by firms depreciate firm value. While Cheung et al. (2009)a, b find that only asset purchases (not sales) reduce the firm value, Ge et al. (2010) show that asset sales (not purchases) destroy the valuations of Chinese firms. Overall, research has focused on studying RPT types based on cash inflow or outflow and finds ambiguous evidence.

Some Indian studies have also examined RPT items based on cash-flow direction and generally report a positive (or null) effect on firm performance. Srinivasan (2013) explores how RPTs affect ROA and finds no impact except that related-party sales reduce ROA. Bansal and Thenmozhi (2020) studied how ownership concentration may affect RPTs and found that founder ownership concentration encourages beneficial RPTs and that RPTs in India enhance firm value. Tripathi et al. (2020) analyze expenses, sales and loans. While expenses increase value, loans given are valuation agnostic. There is little overlap between our study and past Indian studies regarding variables studied, hypotheses and findings.

Only a few studies analyze composite RPTs (Bona-Sánchez et al., 2017; Kang et al., 2014). They assume tunneling is independent of the cash flow direction since firms can buy (sell) at higher (lower) prices to tunnel value from both the buy and sell sides. But even this small set presents mixed evidence about operating and capital asset RPTs. Bona-Sánchez et al. (2017) find that capital and operating RPTs reduce Spanish firms' value. Kang et al. (2014) find that value decline happens only in the five largest chaebols where divergence is very high. Evidence is more explicit that RPLGs reduce value in both developed (Bona-Sánchez et al., 2017; Gordon et al., 2004; Kohlbeck and Mayhew, 2010) and developing markets (Berkman et al., 2009; Jiang et al., 2010).

Literature has focused on RPT items (i.e. selling goods, buying assets, equity assets, other receivables or compensation) suitable to the local setting. These RPTs are sometimes unfit for different institutions and reporting requirements. For example, other receivable is not a typical reported RPT type in India. While locally relevant variables capture newer ways of making specific RPTs, their value effect becomes inconsistent across contexts. Results also differ for an RPT type in the same country, say China (Cheung et al., 2009a, b; Wang et al., 2019; Wong et al., 2015). It suggests that the value effects of RPTs can be institution-dependent. Since propping is more common in China than in other countries, results from China or Hong Kong to other countries are difficult to apply (Bona-Sánchez et al., 2017). Also, past studies have not paid as much attention to composite RPTs as cash inflow/outflow-based measures. Since theory has lagged behind empirical effort (Atanasov et al., 2014), research finds it challenging to consistently identify dubious RPTs across institutional contexts (Bona-Sánchez et al., 2017; Pizzo, 2013).

3.2 The role of counterparties

Another reason that limits the understanding of dubious RPTs is the lack of attention to examining tunneling incentives of counterparties to RPTs. Indian rules require disclosure about not only RPT types but also counterparties. Our classification of counterparties into holding and control types follows the premise of voting-cash-flow rights divergence literature that parties holding greater ownership have higher tunneling incentives (Claessens et al., 2000) than subsidiaries or associates under the control of the firm. This literature suggests that controlling owners have more incentives to tunnel funds from listed firms, where they have lower ownership, to firms held by owners (or relatives) with higher cash flow rights. Research also finds that disproportional ownership increases earnings management (Mindzak and Zeng, 2018). The classification of holding parties (owners/directors/relatives/holding companies) classifies parties who control the firm and have similar rent extraction incentives. Since RPTs with each constituent related-party are infrequent, the classification provides enough RPT frequencies to provide statistical power to models. Also, the classification of controlled parties (subsidiaries and associates) groups “parties under control” with lower controller ownership and low extraction incentives. The controlled parties are generally operating companies in an Indian group. This classification may allow policymakers to frame stricter rules for holding parties and give more latitude to controlled counterparties.

However, past works have not paid sufficient attention to analyzing counterparties, except Kohlbeck and Mayhew (2017). Kohlbeck and Mayhew (2017) classifyies directors, owners, consolidated subsidiaries as DOS, unconsolidated subsidiaries and associates as investees. They also label select RPTs with counterparties as tone and business types. Business RPTs serve business purposes. Kohlbeck and Mayhew (2017) specify loans, borrowings, guarantees, overhead reimbursement, related business and stock deals with investees and leasing with DOS as business RPTs. Loans, borrowings, guarantees, overhead reimbursement and stock deals with DOS are tone RPTs, reflecting insider opportunism. They also consider consulting, legal/investment services/unrelated business between DOS and investees as tone RPTs.

The Kohlbeck and Mayhew (2017) typology is complex. They group several RPT items and collect them as tone or business. This typology rationale of individual RPT items (say, overhead reimbursement or lease, or related business) as tone or business remains unconfirmed in empirical findings. Their study is also too close to the US context. It contains many RPTs (overhead reimbursement, related business and consulting/legal services) attractive for insiders to extract rent. However, these items may not be the most attractive RPTs for family owners to exploit for private benefits. In India, RPLGs, buying/selling assets, real estate, investments and buying and selling goods/services are the main RPT types. Due to group structure with the proliferation of subsidiaries and associates with a relative at the helm, Indian firms rely more on internal capital and product markets to grow or extract rent. Hence, the context is quite different from that of widely held firms.

This paper's definition of counterparties differs significantly from Kohlbeck and Mayhew (2017). Unlike them, I regard relatives of controlling owners or directors as crucial holding parties. They regard consolidated subsidiaries as DOS. However, in line with the divergence literature, I regard subsidiaries as controlled parties. Unlike our study, Kohlbeck and Mayhew (2017) specify that RPLGs can be tone or business, depending on counterparties. It remains unexplained why RPLGs to investees or consulting, legal/investment services/unrelated business between DOS and investees should be tone RPTs. I expect RPLGs to be dubious RPTs irrespective of counterparty types. Rather than labeling some RPTs as tone RPTs ex ante, as in Kohlbeck and Mayhew (2017), I study the interaction of RPT types with counterparties to build and test hypotheses about whether specific RPTs reduce value.

Fooladi and Farhadi (2019) also classify a group of RPTs as beneficial or detrimental based on the value consequences of different RPT types in past studies. The detrimental RPTs are loans/guarantees given, receivables and asset purchases/sales from all parties except subsidiaries. In comparison, beneficial RPTs are loans/guarantees received, receivables and asset purchases/sales with subsidiaries. Clearly, they consider all RPTs with all non-subsidiaries as detrimental and subsidiaries as beneficial (see Table A1, page 224). This paper's classification is more comprehensive than Fooladi and Farhadi's. Unlike studies that decide value relevance ex ante, my study builds cogent hypotheses to delineate what may separate the dubious from benign RPTs.

4. Hypotheses

I rely upon three criteria [2] that broadly characterize “extractability” (i.e. the attractiveness of an RPT type to value extraction) to separate abusive from efficient RPTs. First, dubious RPTs should be valuable, that is, the controller's incentive to tunnel assets is likely related to the size or the yield (or proportion) of financial gains for every dollar extracted (Johnson et al., 2009) vis-à-vis penalty. Second, it should be concealable, that is, the controller will likely design RPTs to reduce the chance of detection. Initially, firms may choose a soft account to record the transaction (Dechow et al., 2011), followed by adjustment entries later to hide it. Third, counterparty type, that is, higher extraction incentives of a counterparty (i.e. holding party), may affect the abuse of RPTs.

4.1 The accounting nature of RPTs and value implications

Operating RPTs are related to operating revenue/expense accounts. These accounts are challenging to manipulate since operating income/expense sources are more specific to routine business, occur frequently, are settled within the same year and are “easy pluck” to be intensively checked by auditors. Operating entries about cash, inputs, wage, leases, royalties, inventory and trade receivables are hard accounts tied to routine matters, and manipulation will likely entail forgery in contracts. Hence, the chance of discovering a manipulation of operating entries is higher. Also, financial gains extracted would be a small part of the transaction (above production costs) to discourage its abuse. The family may not want to approach the board for approval periodically. As such, some studies find adverse value effects of operating RPTs (Bona-Sánchez et al., 2017; Cheung et al., 2006), and a few from Asia-Pacific - mainly China - show that firm value increases with related sales (Wang et al., 2019; Wong et al., 2015) (see Section 3.1). Kohlbeck and Mayhew (2017) also show that operating RPTs are unrelated to restatements in US firms.

In addition, operating RPTs help reduce transaction costs. Given high asset specificity, sunk costs and complexity, and when opportunism or control rights are infeasible or too costly to include in a contract with outsiders, operating RPTs can be an efficient mechanism. These RPTs are more useful as they reduce transaction costs of trading with outsiders, enforce property rights and transfer risks in emerging markets characterized by weaker institutions (Khanna and Palepu, 2000; Khanna and Yafeh, 2007; Fisman and Khanna, 2004). It can reduce operational and unit costs, provide scale economies, release more internal resources in the production process, increase asset utilization by using unused capacity and inventory (Fisman and Khanna, 2004; Khanna and Yafeh, 2007) and allow focus on the core business. Overall, operating RPTs can help reduce operational and overhead expenses and increase asset utilization in family firms.

Past studies have used operating expenses and inefficient asset utilization as measures of agency costs (Ang et al., 2000). Reduced operating expenses reflect improvement in cost discipline, reduction in perks and superior profitability. Similarly, better asset utilization implies investments in positive net present value projects, increased managerial efforts and reduced shirking in utilizing assets (Ang et al., 2000; Fleming et al., 2005). Overall, operational efficiency gains and lower extractability of operating RPTs will positively impact the firm value of family firms. The value effect will likely be positive in a group where several unrelated businesses provide larger and smoother access to the internal capital market (Fisman and Khanna, 2004). Hence. taking the arguments together, I hypothesize that:

H1a.

Operating RPTs will positively affect firm value in family firms.

H1b.

Operating RPTs will positively affect operating efficiency in family firms.

Non-operating RPTs like RPLGs are “soft” accounts (Dechow et al., 2011). These RPTs do not reflect current operations tightly and get adjusted over many years by showing a loss with an explanatory note of excuse, thus decreasing the chance of detection; the longer the adjustment period, the lower the chance of detection of fictitious related-party loans. Research finds that firms with concentrated or family ownership disclose less or inconsistent information (Darmadi and Sodikin, 2013; Fan et al., 2022; Yeung and Lento, 2020) and are less likely to employ high-quality auditors (Darmadi, 2016). Bertrand et al. (2002) find that non-operating RPTs are preferred for tunneling in India as such RPTs allow wide discretion. Also, controllers can extract a much more significant financial gain from even smaller loans. Berkman et al. (2009) find RPLGs behind investor expropriation in China. Jiang et al. (2010) show that owners of Chinese firms tunnel value through intercorporate loans. Research also asserts that firms often do not pay the principal or the interest on loans received from related parties (Berkman et al., 2009; Kali and Sarkar, 2011). In sum, high extractability and past evidence on adverse value effects (Bona-Sánchez et al., 2017; Claessens et al., 2006; Jiang et al., 2010) suggest that RPLGs will reduce firm value in India. Hence, I hypothesize that:

H2.

Loan/Guarantee RPTs will negatively affect firm value in family firms.

Capital asset RPTs may help overcome institutional void (Khanna and Palepu, 2000) and impart scale/scope economies by providing critical assets (Fisman and Khanna, 2004). However, these RPTs are quite infrequent, do not occur ordinarily and involve larger volumes than operating RPTs. Buying/selling capital assets can be over- (under)-invoiced by related parties to siphon money out. The amount not coming back is moved to receivables to evade detection. This way, the amount extracted can be a significant fraction of RPTs. Firms hold assets for several years, and market prices may be unavailable at the settlement time when losses are recognized. Hence, asset RPTs involve more discretion than operating RPTs and may reduce the firm value (Bona-Sánchez et al., 2017; Ge et al., 2010; Kang et al., 2014). Bae et al. (2002) show intra-group acquisitions in South Korea expropriate value. Thus, it seems capital asset RPTs may raise expropriation concerns among investors. Hence, I hypothesize that:

H3.

Capital asset RPTs will negatively affect firm value in family firms.

4.2 Counterparty types and the expropriation logic

Besides the ad hoc selection of RPT types, past evidence has remained mixed due to the lack of attention to RPT counterparties. Since upper-level parties carry sizeable tunneling incentives, owners tunnel value from the bottom level, where ownership is low, to the upper level, where ownership is high (Johnson et al., 2000). Since incentives are alike, I aggregate each RPT type with directors, owners, relatives and holding firms parties close to the owner. It is from these parties the owner can siphon the maximum amount home. In contrast, the extracted private share from controlled parties will likely be lower and depend upon the ease of evasion.

Kohlbeck and Mayhew (2010) suggest that RPLGs to directors are the worst diversions of assets. Kohlbeck and Mayhew (2017) show that “tone at the top” RPTs signal a high risk of financial misstatements. The Companies Act 1956 of India does not prohibit RPLGs with holding parties. These loans also receive credit support or guarantees backed by other holding parties. Given the chance of financial misstatements and the extraction incentives of holding parties to gain from RPLGs, it is likely that these RPTs with holding parties may lead to expropriation. Due to their sparing use, lower chance of detection and higher extraction yield, RPLGs to controlled parties remain an attractive way to tunnel value. While RPLGs can be tempting for quick and genuine fund transfers, the extraction opportunity seems too strong to resist. Firms ostensibly pump capital into subsidiaries and associates helmed by relatives to finance expansion. Given the high extractability of RPLGs, firms can divert funds through subsidiaries and associates. Using shell companies can make fund transfer and ownership trails extremely difficult to unravel. Hence, I hypothesize that:

H4.

Loan/guarantee RPTs to holding or controlled parties will negatively affect firm value in family firms.

While building hypothesis H1, I argued that recurring nature, easier tracking, limited extraction incentives and transaction cost savings make operating RPTs more beneficial for firms. Still, operating RPTs with holding parties may raise concerns about conflict of interest, if not outright value extraction. Despite lower yield and smaller size (INR 0.12 billion, Tabel A1), holding parties have incentives to trade services/goods at better pricing to make money. Individual directors or relatives may find even the small size attractive (Gozlugol, 2021), undertake phantom RPTs [3] and can easily conceal them. Also, they are typically not operating entities and hence lack capabilities in minimizing transaction costs and providing scale/scope advantages in trading goods/services. However, legal and entrepreneurial services (e.g. expenses in procuring government licenses, contracts and political connections) by holding parties may appear beneficial to firms.

Regarding operating RPTs with controlled parties, the amount extracted by the controller will likely be small due to the higher costs involved in routing the money through the chain. Also, given the higher visibility of operating RPTs, the chance of detection and the consequent loss of reputation for a small financial gain is high. Besides, controllers have lower abilities and incentives to siphon funds via controlled parties since non-controllers have significant ownership in associate firms. As operating companies in a group in a specific domain (e.g. construction), the controlled parties can improve scale/scope economies to reduce operating costs and increase asset utilization. Thus, operating RPTs with controlled parties are more likely to be efficient transactions. Hence, I hypothesize that:

H5a.

Operating RPTs with controlled parties will positively affect firm value in family firms.

H5b.

Operating RPTs with holding parties will negatively affect firm value in family firms.

Appendix 1 provides the direction of expected relationships.

5. Data, variables and methodology

5.1 Sample and data

I use the Prowess database to collect stock prices, balance sheets, RPTs and ownership data from 2001–11, before the new Companies Act came in 2013. Section 185 of the new Act severely curtails RPTs, especially loans/guarantees. The prior period allows us to examine firms' behavior toward RPTs in the nascent stage of corporate governance reforms. It also allows us to diagnose how firms' approached RPTs (and value effects) when markets, investors and regulators were less informed about RPTs and paid less attention to them. No past Indian study, except for Srinivasan (2013), has studied RPTs in this early period of transplant of the US/UK governance regime. The study will help assess the direction of the Company's Act 2013 regarding RPTs. Several studies published in reputed journals have used Prowess (Bertrand et al., 2002; Khanna and Palepu, 2000; Gopalan et al., 2007, among others). The disclosure of the controller's identity follows clauses 35 (2001–2004), clause 49 (since 2004) and the takeover code. I include firms with more than ₹50 million market cap, sort them by fiscal 2001 market cap and exclude all government firms, financial firms and firms lacking regular data. Finally, the sample comprises 578 BSE-listed firms.

5.2 Related party transactions

Indian Accounting Standard and clause 49 of listing requirements mandate disclosure of RPTs. The RPTs disclosed are receipts (or payments) involved in the buy/sale of goods, services or assets; leasing property, vice versa; loans, guarantees and license fees. The related parties disclosed by Indian regulation are (1) associates under control; (2) subsidiaries; (3) key people like directors, senior executives and major owners; (4) relatives of key people; (5) holding companies. The study defines holding parties as composing controlling owners, directors, top managers, relatives and holding companies. The controlled parties comprise subsidiaries with over 50% ownership, joint ventures and associates under control with ownership between 20–50%. I collect data on the following RPTs: (1) total RPTs with all parties, (2) buy/sell, capital asset and loan/guarantees with all parties, holding and controlled parties. Indian firms' widespread use of RPTs at more than 85% of firm years makes my use of RPT volume a robust measure.

5.3 Identifying the controlling owner or family

Indian regulation requires firms to identify promoters (controllers) and disclose the ownership of all major shareholders. I identified family members from major owners disclosed. Family members mainly have the same family/community surname. A specified family holds at least 20% ownership - individually, via trust or firms - in a family firm (Anderson and Reeb, 2003). I measure ownership concentration as the aggregate equity shareholding of a firm's controlling family.

5.4 Variables

The study uses Tobin's q, a measure of long-term firm value, as the main dependent variable. I measure it as the sum of market cap and liabilities over total assets.

The key explanatory variables are RPT: the ratio of the total RPTs over sales; GS: buy/sell of goods/services over sales; LG: annual related-party loans/guarantees over assets; CA: related party buy/sell of capital assets over sales; CGS: operating RPTs with controlled parties over sales; CA: capital asset RPTs over sales; HCA: capital asset RPTs with holding parties over sales; CCA: capital asset RPTs with controlled parties over sales; HLG: loans/guarantees to holding parties; CLG: loans/guarantees to controlled parties; FAM denotes a family firm, zero otherwise.

The control variables are: Far: fixed assets over assets; Lev: total debt over assets; Size: log of total assets; Sta: sales over assets; Age: log of firm-age since incorporation; Vol: volatility of one-year weekly stock returns; Capx: capital expenditure over assets. Leverage captures the effects of capital structure on firm value; debt can mitigate (or even distort) agency conflict. Volatility captures the relation between risk and value; fixed assets and capital expenditure capture the ease (or the difficulty) of monitoring discretionary spending, impacting firm value. Firm age and size capture the operating and contracting environment. Larger or older firms may benefit from scale economies but also incur high monitoring costs. This way, I include control variables per the ownership-value literature and control for alternative effects on firm value. I also include return on assets (Roa) measured as EBITDA/assets ratio. ROA is expected to affect firm value positively. I also use year and industry dummies (India's two-digit industry code equivalent of the SIC code) to control time and industry effects that may affect the market cap of firms.

5.5 Summary statistics

Family firms comprise 65%, and group firms 45% of this sample of 578 firms. Table A3 shows that family firms have lower q (1.33), higher total assets, more leverage (35%), fewer profits (12%) and higher fixed assets (32%), and ensure better asset utilization than non-family firms.

Table A2 shows the frequency of RPT types. Buy/Sell of goods and services (GS) is the most frequent RPT type (80% firm-years), followed by capital asset RPTs (CA) (about 38%) and RPLGs in a year (LG) (12.10%). Operating RPTs with controlled parties (CGS) are frequent (64.6%), followed by capital asset RPTs with controlled parties (CCA) (36.32%). Operating RPTs with holding parties (HGS) are also quite common (69.7%). However, capital asset RPTs and RPLGs to holding parties are highly infrequent RPT types (3.7 and 1%).

Operating RPTs form the largest aggregate amount (Rs. 7,950bn) in the study period, followed by RPLGs (Rs. 3,778bn) and capital assets (Rs. 3,039bn). RPLGs are the largest amount per firm year (Rs. 6.73bn), followed by operating RPTs (Rs. 2.15bn) and capital assets RPTs (Rs. 1.84bn). Operating and capital assets RPTs with controlled parties comprise 95 and 96% of respective RPT types.

Table A3 shows the mean and volatility of variables. RPLGs (LG) is 1% of assets. The t-test indicates that LG by family firms to all related parties, mainly controlled parties, is larger than LG by non-family firms. The mean operating (GS) and capital assets (CA) RPTs are 28 and 27% of sales. Operating RPTs by family firms are larger than those by non-family firms. Capital RPTs with all parties (also holding parties) do not differ across family and non-family firms. Family firms undertake larger capital asset RPTs with controlled parties than non-family firms. Also, total RPTs by firms are 12% of total firm assets and do not differ across family and non-family firms.

6. Results

6.1 Value effects of RPTs

6.1.1 RPT types and value effects

In Table 1, I use fixed effects (FE) regressions specified in equation (1) to analyze whether RPT volume is related to value expropriation in family firms. βt,βind and βi are year, industry and firm effects. RPT∼ is an array of current and lagged RPT types capturing shorter- and longer-term effects. Also, lagging variables is a standard research method to mitigate potential reverse causality. I also checked multicollinearity (VIF<2) for each current and lagged RPT variable.

(1)ln(Tobq) =α0 +β1RPT∼ +β2Far +β3Lev +β4 Sta +β5 Size +β6 Age +β7 Vol +β8 Capx +β9 Roa +β10 Growth +βi+βt +βind +εit

Model (1) shows that a unit increase in loans/guarantees by family firms in the previous year reduces the firm value by 27% (p < 0.05). Also, a unit increase in loans/guarantees made two years ago reduces the firm value by 42% (p ≤ 0.05). The result indicates a longer-term effect of loans/guarantees on value destruction [4]. Model (2) uses the log-log form regression [5] and finds that firm value decreases by 1.5%, with a 1% rise in average capital asset RPTs over the last two years (p = 0.06). Capital RPTs in the current year are insignificant. Since operating RPTs are highly frequent and get settled within the same year, I regress the log of Tobin's q on current and lagged average operating RPTs made over the last two years. Model (3) shows that a unit increase in operating RPTs increases firm value by 0.5% (p ≤ 0.05). A log-log form regression (unreported) shows that a 1% rise in operating RPTs leads to a 1.7% increase in family firm value (p ≤ 0.05). VIF between current and lagged RPT variables is less than two in all models.

6.1.2 RPTs with controlled or holding parties

In Table 2, I study RPTs with counterparties. All models use the fixed-effects models (equation 1) with industry and year effects. I use the last two years' averages to capture long-term effects.

Model (1) shows that firm value declines by 85% for a unit rise in two-year lagged (p ≤ 0.01) and by 119% for one-year lagged RPLGs to holding parties (p ≤ 0.01). Model (2) shows that firm value declines by 47% for a unit increase in two-year lagged (p ≤ 0.05) and by 24% for one-year lagged RPLGs to controlled parties (p ≤ 0.10). Both models (1) and (2) support hypothesis H4.

Capital RPTs happen infrequently but tend to correlate across two consecutive years. As such, successive capital asset RPTs create large multicollinearity in models. Hence, to capture long-term effects yet mitigate multicollinearity, I introduce the last two years' average of capital asset RPTs. VIF continues to be under 2. Capital asset RPTs with holding parties are a highly infrequent RPT type in a panel. I avoid the log of CA_t with holding parties in model (3) since it drops most observations and weakens the panel data analysis. Regressing log q, I find that a unit increase in past capital asset RPTs with holding parties increases firm value by 0.5% (p ≤ 0.01). The current capital asset RPTs are insignificant and are positively related to firm value when lagged RPTs are absent in the model. Following model (2) (Table 1) strategy, I use the log-log form in model (4) in Table 2 since capital asset RPTs with controlled parties are frequent and large in some years. Log form reduces this dispersion and provides the best fit. Model (4) shows that a 1% increase in average capital asset RPTs over the last two years (log) with controlled parties reduces the firm value by 2% (p ≤ 0.05). The value effect of current capital asset RPTs (log) is insignificant.

I regress the log of q on current and lagged average operating RPTs over the last two years. Model (5) shows that firm value increases by 0.6% for a unit increase in current operating RPTs with holding parties (p ≤ 0.01). However, hypothesis 5a finds support as firm value declines by 1.6%, showing a detrimental longer-term effect (p ≤ 0.01). Model (6) shows that operating RPTs with controlled parties bears a null impact on firm value in both the long term and short term.

6.1.3 Operating RPTs and operating performance

In Table 3, I use fixed-effects models (equation (2)) to evaluate the impact of current and lagged operating RPTs on profits (EBITDA/assets), asset utilization (sales/assets) and expenses.

(2)Efficiency =α0 +β1GS +β2Far +β3Lev +β4 Sta +β5 Size +β6 Age +β7 Vol +β8 Capx +βi+βt +βind +εit

Model (1) shows that profitability (EBITDA/assets) rises with two-year lagged operating RPTs by family firms (p ≤ 0.01). Model (2) shows that asset utilization rises with two-year lagged operating RPTs (p ≤ 0.05); current and one-year lagged operating-RPTs are insignificant. Model (3) shows that while the increase in asset utilization rises with operating RPTs with controlled parties by 0.018 units (p ≤ 0.05), it reduces by 0.08 units when family firms increase operating RPTs with holding parties. Related party reimbursements happen when a related supplier incurs expenditure on behalf of the recipient. These could be fees, taxes, logistics charges, advertising, packing, commission, traveling, agency arrangements and sales expenses. Regression of SGA expenses (log) over assets in model (4) shows that a unit rise in operating RPTs reduces SGA expenses by 0.01 units (p ≤ 0.05). Likewise, the regression of operating expenses (log) over assets in model (5) shows a rise in current operating RPTs reduces operating expenses by 3.3% (p ≤ 0.01). Overall, the results strongly support the efficiency hypothesis (Ryngaert and Thomas, 2012).

6.2 Family firms and RPT behavior

Extant research does not find evidence and merely conjectures that family firms make more RPTs than non-family firms. It does not record the family firms' preferences for specific RPT types. However, reducing transaction costs by operating RPTs can generate new firm-specific advantages, facilitate coordination and lessen opportunism and imperfect effort by non-family partners. Hence, family firms are more likely to internalize operating RPTs than non-family firms.

Table 4 examines family firm behavior toward specific RPT types in three ways: first, whether family firms undertake more sizeable RPTs; second, whether they are more likely to use some RPT types; and third, whether they are more likely to cross the materiality threshold. The threshold provides another measure to assess family firms' tendency to undertake voluminous RPTs in a year. As per the Companies Act 1956, the materiality threshold for loans/guarantees is 30% of paid-up capital and dividend paid and 10% of the turnover for operating RPTs.

Model (1) shows that family firms lend more voluminous RPLGs than non-family firms (p ≤ 0.05). However, model (2) predicts that family firms are less likely to use RPLGs than non-family firms (p ≤ 0.10). Also, model (3) shows that family firms are more likely to cross the RPLG materiality threshold (p ≤ 0.05). I do not find a tendency of family firms vis-à-vis non-family firms to use RPLGs with holding parties. The findings suggest that family firms are less likely to undertake RPLGs but advance big-ticket RPLGs than non-family firms.

Next, I analyze the family firm propensity to use capital RPTs. Model (4) shows that family firms use more capital assets RPT volumes than non-family firms (p ≤ 0.10). They are also more likely to use capital RPTs with holding parties than non-family firms (model (5)) (p ≤ 0.10). However, they have a similar propensity as non-family firms to undertake capital asset RPTs.

Finally, I analyze the family firm propensity to use operating-RPTs. Model (6) shows that both family and non-family firms use almost equal operating RPTs volumes. However, model (7) predicts that family firms are less likely to use operating RPTs with holding parties (p ≤ 0.01) and are less likely to cross the materiality threshold than non-family firms (p ≤ 0.10) (Model (8)).

6.2.1 Family firm penchant for making select RPTs: evidence from a natural experiment

I use the background of the Satyam scam [6] and the resulting resignation of independent directors (IDs) from many firms to draw inferences about the association of corporate governance with abusive RPTs. An exogenous shock spurred the exits offers a unique setting to set up a DID model to alleviate unobserved factors. Since the scam first came to notice on December 16, 2008, when the board approved RPTs with related family parties, I keep FY 2008–09 as the year of shock resulting in ID exits. All firms with ID resignations represent the treated, and the other firms that did not experience ID exit as control. This sample recorded eighty ID resignations in 2009. I code 2010 and 2011 as one, and the years before 2009 as zero. Based on this formulation, I specify the DID model in equation (3) as follows:

(3)RPTit =β0 +β1After i +β2 After i x Treat t +β3 X +βi +βt+βind +εit

where After _i equals one if an ID has resigned from firm i in 2009; Treat _t equals 1 for years after 2009, and 0 for 2004 to 2008. βi, βt and βind are firm, year and industry effects.

Table 5 examines whether governance shock from the Satyam scam affects RPTs by firms. Model (1) shows no change in the family firm's propensity to undertake RPLGs before and after the scam. However, model (2) shows a significant positive effect on RPLGs to holding parties (p ≤ 0.01). Model (3) shows that capital asset RPTs reduce significantly after IDs' exit (p ≤ 0.01), especially capital asset RPTs with holding parties (model (4)) (p ≤ 0.01). Model (5) shows that family firms reduce operating RPTs after IDs' exit (p ≤ 0.05). The impact stays significant mainly in the case of operating RPTs with holding parties (model (6)) (p ≤ 0.05).

Overall, I find that family firms continue with dubious RPLGs. However, they slash capital assets and operating RPTs, typified by higher visibility and lower private share.

6.3 Robustness

6.3.1 Comparing family with non-family firms

Table 6 compares the value effects of select RPTs by family firms with non-family firms in panel (A), and family business groups with stand-alone firms in panel (B). A group comprises at least two firms under common family ownership. Model (1) shows that a unit increase in historical capital asset RPTs with holding parties increases the firm value of family firms by 1.8% compared to non-family firms (p ≤ 0.05). Model (2) shows that a unit increase in operating RPTs with all parties increases the value of family firms by 8.2% (p ≤ 0.01). Model (3) finds that an increase in operating RPTs with holding parties increases the value of family firms by 12.6% compared to non-family firms (p ≤ 0.01). Models (4) to (6) show similar positive value effects of capital asset and operating RPTs findings for family groups versus stand-alone firms. Similar regressions (unreported) on RPLGs show no differential effects across family versus non-family or family group versus stand-alone firms. It seems RPLGs are perceived to be equally worse for all firms. Overall, the findings of Table 6 corroborate the results of Table 2.

I also employ a two-step system GMM to examine RPT types. It provides a better way to handle dynamic endogeneity from past firm values and current RPTs. I cap lags to three periods (3 3). The older lags are likely to be exogenous and valid. The exogenous instruments are firm-age, the volatility of returns, year and industry effects. Overall, the models support Table 3 results.

In alternative models, I also evaluate the value effects of the frequency of specific RPTs and their occurrence with counterparty types. I find that results corroborate earlier findings. I use a robust White-estimator, insensitive to adding/removing extreme observations. The models are coarse to covariate selection. Other Tobin's q measures, like the market-to-book and market-to-assets ratios, give similar results. I experimented with a different definition of a family firm that includes entrepreneurial founder-led firms with similar results.

6.4 Summary and discussion

Research has focused on exploring the value relevance of buy- or sale-type RPTs. Very few scholars have studied basic RPT types to better gauge the different drivers affecting their usage. Research has also paid scant attention to tunneling incentives of counterparties. This study analyzes a family firm's choice of RPT types and examines their value impacts to identify value-destroying RPTs.

The finding of value detraction from RPLGs supports past results in different institutional settings (Berkman et al., 2009; Bona-Sánchez et al., 2017; Jiang et al., 2010). I provide the first evidence that family firms undertake larger RPLG volumes less frequently than non-family firms. The value destruction happens irrespective of counterparty types. Notably, RPLGs by both family and non-family firms are equally value-destroying. Due to large yield, easy concealment and movement across group firms or shell companies, RPLGs remain highly extractable RPT type.

Research has examined capital assets and operating RPTs by bifurcating them on cash flow direction, arguing that outflows indicate more value extraction than inflows, with mixed evidence (Cheung et al., 2009a, b; Cheung et al., 2006; Ge et al., 2010; Wang et al., 2019; Wong et al., 2015). I cumulate the buy and sale sides of RPTs since extraction can happen at either end. Very few papers have used this approach (Bona-Sánchez et al., 2017; Kang et al., 2014). The study confirms that capital asset RPTs are value-destroying, as found in Spain (Bona-Sánchez et al., 2017) and South Korea (Kang et al., 2014). I also provide the first evidence that family firms undertake larger capital asset RPTs and create more value than non-family firms. Capital asset RPTs with holding parties are related to higher firm value as family owners use this route to bring valuable assets.

Contrary to past studies (Bona-Sánchez et al., 2017; Cheung et al., 2009a, b), operating RPTs enhance firm value per the efficient transaction view (Ryngaert and Thomas, 2012). These RPTs increase asset utilization and discretionary expenses. In a piece of novel evidence, the gains happen with controlled parties (usually operating companies) with a higher ability to transfer scale/scope economies than holding parties. So far, research has underexplored operating RPTs' effects on operational gains. Operating RPTs create more value in family (group) firms than other firms, indicating significant internal capital and product market efficacy in family (group) firms.

The results convey that neither RPTs' nature nor counterparty incentives alone sufficiently predict dubious RPTs. First, the higher extractability of RPLGs dominates the lower extraction incentives of controlled parties since even subsidiaries can be a vital conduit to siphon loans via opaque shell companies. Second, the holding parties' supplying assets to provide a growth engine dominates higher extraction incentives of holding parties. Third, the large operational efficiency gains come from operating RPTs with controlled parties, generally operating companies in family groups. The controlled parties contribute to reducing operating expenses and holding parties in reducing SGAs. Overall, my classification of counterparties yields some interesting findings.

Literature always conjectured but failed to show that family firms are more likely to use RPTs than non-family firms. They undertake sizeable RPLGs and capital asset RPTs but smaller operating RPT volumes than non-family firms. They also undertake smaller total RPTs than non-family firms, challenging the long-held conjecture about family firms' unique dependence on RPTs. Using the historic Satyam scam context, I show that while family owners reduce valuable operating RPTs, yet continue advancing RPLGs, despite RPTs being in regulators' cross-hairs. Family firms' rigidity toward RPLGs implies that RPT choice depends on their family-specific purposes. Further, capital RPTs present a gray area due to concurrent extraction and benefits. Their sizes, structuring and lack of liquid market to determine fair prices increase extractability. However, lower yield, ease of physical verification, tax rules and paper trails make capital RPTs less suitable for extraction. It is harder to undertake phantom asset RPTs. Also, the capital infusion can be a source of value for family firms in high-growth opportunity markets (Durnev and Kim, 2005) like India.

While family firms favor some RPTs types to extract value and some to create value, research gives little insight into family attributes behind this schismatic behavior of family firms. Multi-generational family firms with small ownership, outsider CEOs and reputation commitment may behave responsibly. Future studies can analyze family attributes that affect divergent uses of RPTs. Also, since institutions can affect the frequency, evasion and extractability of select RPTs, future studies can give more attention to sources of operating RPTs' benefits in different institutional contexts. It has not been attempted so far in research. Finally, the role of counterparty incentives in capital assets and operating RPTs needs more examination in other settings.

The findings point to some crucial issues regarding RPTs. First, the study does not single out family firms for dubious use of all RPTs. Non-family firms are also active participants, although they use dubious RPTs in much smaller sizes. Second, investors, auditors or creditors must pay closer attention to RPLGs. RPLGs are highly suitable for siphoning funds using shell firms and subsidiaries. As such, the government has sought to deregister shell companies and disallow multi-step-down subsidiaries. Third, although operating and capital RPTs with holding parties seem beneficial, varying counterparty incentives suggest that solid procedural safeguards for arm's length contracting and insider accountability are necessary. Auditors should scrutinize asset RPTs intensively as they are a gray mechanism with benefits and extraction opportunities. Although operating RPTs show minimal extraction, stringent rules on RPLGs may push family firms to siphon funds through them. Strengthening physical verification in audits may deter phantom RPTs.

Overall, these results may help assess the later direction of reforms and clarify the dilemma Indian regulators face in balancing the abusive and business sides of RPTs.

7. Conclusion

I argue that RPTs' misuse hinges on value extractability and counterparty incentives. Operating RPTs seem efficient contracts, but RPLGs and capital RPTs reduce value. The transaction attributes of RPTs interact with counterparty types to produce divergent motives about RPTs qualifying tone-at-the-top view and the divergence literature. RPLGs remain extremely extractable by all counterparties. However, family firms benefit more from capital asset transactions with holding parties and goods/services trading with controlled parties than non-family firms. Yet, dubious RPTs seem more integral to family firms' preferences than non-family firms. They make higher volumes of dubious RPTs and smaller sizes of operating RPTs than non-family firms. Furthermore, they can reduce more visible and valuable operating RPTs but may persist in RPLGs.

Table 1

The effect of volumes of RPT types on firm value in family firms

	Fixed effects regression of firm value Log of Tobin's Q
	(1)	(2)	(3)
LGt	−0.128
	(0.149)
LGt-1	−0.271**
	(0.132)
LGt-2	−0.421**
	(0.177)
CAt		0.007
		(0.007)
Avg CAt-1,t-2		−0.015*
		(0.008)
GSt			0.005**
			(0.002)
Avg GSt-1,t-2			−0.002
			(0.003)
Far	−0.548***	−0.891***	−0.525***
	(0.199)	(0.271)	(0.198)
Lev	0.315***	0.328	0.312***
	(0.039)	(0.217)	(0.038)
Sta	0.062	−0.077	0.072*
	(0.040)	(0.179)	(0.038)
Vol	−0.005	0.039	−0.012
	(0.033)	(0.050)	(0.031)
Size	−0.255***	−0.328***	−0.253***
	(0.055)	(0.070)	(0.056)
Age	0.503	−0.004	0.394
	(0.333)	(0.364)	(0.318)
Capx	−0.006	0.506*	0.025
	(0.099)	(0.267)	(0.089)
Roa	0.784***	1.786***	0.805***
	(0.283)	(0.475)	(0.287)
Growth	−0.000***	−0.000	0.000*
	(0.000)	(0.000)	(0.000)
Firm effects	Yes	Yes	Yes
Year effects	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes
Constant	0.826	3.167**	1.148
	(0.990)	(1.220)	(0.962)
No. of Obs	2,207	837	2,210
R-squared	0.46	0.52	0.46
No. of firms	376	226	376
Prob(F)	0.00	0.00	0.00

Note(s): Table 1 estimates fixed effects regressions of the firm value on lagged RPT type variables. Table A3 defines various RPT types and firm variables. Model (2) uses the log-log form to control the dispersion of large CA values. The average VIF of RPT variables is less than 2 in all models. *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table 2

Effect of RPT volume with related counterparty types on the firm value

Fixed effects regression of the log of Tobin's Q
Counterparty types	Holding	Controlled	Holding	Controlled	Holding	Controlled
	(1)	(2)	(3)	(4)	(5)	(6)
LGt	0.195	−0.269
	(0.127)	(0.216)
LGt-1	−1.193***	−0.240*
	(0.272)	(0.141)
LGt-2	−0.850***	−0.467**
	(0.238)	(0.205)
CAt			0.002	0.006
			(0.001)	(0.007)
Avg CAt-1,t-2			0.005***	−0.020**
			(0.001)	(0.008)
GSt					0.006***	0.007
					(0.001)	(0.006)
Avg GSt-1,t−2					−0.016***	0.002
					(0.003)	(0.004)
Far	−0.327**	−0.397***	−0.459***	−0.850***	−0.527***	−0.528***
	(0.148)	(0.138)	(0.171)	(0.297)	(0.198)	(0.199)
Lev	0.360***	0.371***	0.345***	0.352	0.316***	0.312***
	(0.101)	(0.101)	(0.059)	(0.229)	(0.041)	(0.038)
Sta	0.106**	0.149***	0.049	−0.086	0.063	0.072*
	(0.043)	(0.038)	(0.042)	(0.178)	(0.040)	(0.038)
Vol	−0.066**	−0.065**	−0.035	0.038	−0.017	−0.014
	(0.026)	(0.027)	(0.027)	(0.050)	(0.030)	(0.031)
Size	−0.027	0.010	−0.150***	−0.335***	−0.257***	−0.256***
	(0.045)	(0.040)	(0.050)	(0.076)	(0.055)	(0.056)
Age	−0.226	−0.459**	0.319	−0.086	0.417	0.389
	(0.258)	(0.224)	(0.275)	(0.416)	(0.320)	(0.317)
Capx	0.080	0.139	−0.038	0.511*	−0.001	0.026
	(0.112)	(0.105)	(0.087)	(0.267)	(0.098)	(0.090)
Roa	0.603**	0.185	0.883***	1.795***	0.808***	0.802***
	(0.288)	(0.248)	(0.299)	(0.488)	(0.285)	(0.287)
Growth	−0.000	−0.000*	0.000	−0.000	0.000**	0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Firm effects	Yes	Yes	Yes	Yes	Yes	Yes
Year effects	Yes	Yes	Yes	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes	Yes	Yes	Yes
Constant	0.414	0.820	0.245	3.461**	1.121	1.192
	(0.688)	(0.620)	(0.807)	(1.376)	(0.967)	(0.957)
No. of obs	3,235	3,562	2,563	802	2,210	2,210
R-squared	0.36	0.36	0.40	0.52	0.46	0.46
No. of firms	377	377	377	218	376	376
Prob(F)	0.00	0.00	0.00	0.00	0.00	0.00

Note(s): Table 2 shows fixed effects regressions of Tobin's q (log) on lagged RPT type variables. Table A3 defines various RPT types and firm variables. CA_t and Avg CA_t are in the log form in model (4). Model (3) avoids the log since CA_t with holding parties is highly infrequent. VIF of RPT variables is less than 2 in all models. Robust standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table 3

Effect of operating RPT volume on operating efficiency

Variables	Return on assets	Asset utilization	Inc. In asset utilization	Sales and general expenses (SGAs)	Total operating expenses
	(1)	(2)	(3)	(4)	(5)
GSt	−0.001	−0.008		−0.001**	−0.032***
	(0.001)	(0.006)		(0.001)	(0.008)
GSt-1	0.002	0.007		−0.000	−0.007
	(0.002)	(0.007)		(0.001)	(0.015)
GSt−2	0.002***	0.007**		−0.001	0.010
	(0.001)	(0.003)		(0.001)	(0.010)
HGSt			−0.067
			(0.067)
HGSt−1			0.077*
			(0.043)
HGSt−2			−0.085*
			(0.046)
CGSt			0.002
			(0.003)
CGSt−1			−0.003
			(0.007)
CGSt-2			0.018**
			(0.009)
Far	0.048	0.151	0.334***	0.178***	1.318***
	(0.034)	(0.311)	(0.091)	(0.041)	(0.434)
Lev	−0.025*	0.031	0.084	−0.127***	−0.139***
	(0.014)	(0.108)	(0.074)	(0.042)	(0.038)
Size	−0.010	−0.303*	−0.099***	−0.048***	−0.145
	(0.010)	(0.158)	(0.027)	(0.010)	(0.116)
Age	−0.041	0.204	0.233*	0.188***	0.301
	(0.040)	(0.157)	(0.131)	(0.070)	(0.357)
lnTobq	0.049***	0.124*	0.018	−0.006	0.166***
	(0.010)	(0.064)	(0.021)	(0.010)	(0.060)
Firm effects	Yes	Yes	Yes	Yes	Yes
Year effects	Yes	Yes	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes	Yes	Yes
Constant	0.319**	2.773*	−0.022	0.049	−0.654
	(0.136)	(1.417)	(0.506)	(0.259)	(1.531)
No. of obs	2,575	2,575	2,161	2,161	2,575
No. of firms	379	379	311	311	379
Prob(F)	0.00	0.00	0.00	0.00	0.00

Note(s): Table 3 shows fixed effects regressions of operating performance on two-year lagged operating RPTs. Table A3 defines various RPT types and firm variables. The average VIF of RPT variables is less than 2. *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table 4

Family influence on RPT volume and disclosure

Dep. Variables	Loans/Guarantees			Capital assets		Operating RPTs
	Volume	Disclosure	Materiality disclosure	Volume	Disclosure	Volume	Disclosure	Materiality disclosure
		All	All		Holding		Holding	All
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
FAM	0.004**	−0.220*	0.372**	0.409*	0.438*	0.282	−0.280***	−0.149*
	(0.002)	(0.122)	(0.188)	(0.241)	(0.233)	(0.217)	(0.102)	(0.086)
Tobq	−0.000	−0.060	0.031	0.237*	0.209*	0.166	0.025	0.161***
	(0.001)	(0.072)	(0.098)	(0.137)	(0.108)	(0.146)	(0.061)	(0.052)
Far	−0.011**	−0.710**	−1.791***	−0.435**	0.165	−0.296	0.009	−1.360***
	(0.005)	(0.310)	(0.571)	(0.218)	(0.513)	(0.408)	(0.232)	(0.250)
Lev	0.002	−0.175	0.133	−0.064	−0.028	0.180	−0.316***	0.122
	(0.001)	(0.130)	(0.105)	(0.089)	(0.134)	(0.234)	(0.120)	(0.087)
Size	0.004***	0.390***	0.525***	0.047	0.212***	0.020	0.242***	0.100***
	(0.002)	(0.036)	(0.047)	(0.045)	(0.051)	(0.056)	(0.032)	(0.025)
Age	−0.004	−0.212**	−0.152	−0.102	0.048	−0.110	0.045	−0.287***
	(0.003)	(0.092)	(0.124)	(0.162)	(0.173)	(0.103)	(0.083)	(0.070)
Vol	0.002	−0.104**	−0.145**	0.040	−0.041	0.102	0.010	−0.017
	(0.002)	(0.051)	(0.073)	(0.058)	(0.075)	(0.121)	(0.030)	(0.030)
Capx	−0.000	−0.586	0.293	−0.021	−0.413	−1.797	1.260**	−0.579
	(0.001)	(0.399)	(1.310)	(0.130)	(0.368)	(1.868)	(0.503)	(0.631)
Year effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Reg. model	RE	Logit	Logit	RE	Logit	RE	Logit	Logit
Constant	−0.030***	−5.930***	−9.327***	−0.219	−5.052***	0.051	1.193*	−0.138
	(0.008)	(1.128)	(1.211)	(0.689)	(0.833)	(0.375)	(0.695)	(0.381)
Observations	5,927	3,920	3,920	4,772	3,920	4,373	3,936	3,936
No of firms	571			559		509
Prob > F	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00

Note(s): Table 4 analyzes family firms' behavior toward RPT types vis-à-vis non-family firms. Table A3 defines various RPT types and firm variables. *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table 5

Effect of corporate governance shock on RPT volume

Dependent Variables counterparty types	Loans and guarantees	Loans and guarantees Holding parties	Capital asset RPTs	Capital RPTs Holding parties	Operating RPTs	Operating RPTs Holding parties
	(1)	(2)	(3)	(4)	(5)	(6)
After*Treated	0.006	0.006***	−1.380***	−0.317**	−0.562**	−0.244**
	(0.005)	(0.002)	(0.335)	(0.144)	(0.275)	(0.110)
After	0.018***	0.001	0.601	0.318*	0.991***	0.319**
	(0.006)	(0.003)	(0.406)	(0.174)	(0.334)	(0.134)
Far	−0.025**	−0.002	0.327	0.083	−0.778	−0.001
	(0.012)	(0.005)	(0.795)	(0.341)	(0.653)	(0.262)
Lev	0.005	−0.001	−0.066	−0.045	0.381**	−0.040
	(0.003)	(0.002)	(0.232)	(0.100)	(0.191)	(0.077)
Sta	−0.001	−0.001	−0.321*	−0.054	−1.108***	−0.096*
	(0.003)	(0.001)	(0.174)	(0.075)	(0.143)	(0.057)
Size	0.008***	−0.000	−0.312*	−0.175**	−0.610***	−0.189***
	(0.003)	(0.001)	(0.170)	(0.073)	(0.140)	(0.056)
Age	−0.035**	0.003	0.220	−0.655	−1.109	−0.549*
	(0.015)	(0.007)	(1.012)	(0.434)	(0.831)	(0.333)
Tobq	−0.003	0.001	0.350**	0.083	0.206	0.078
	(0.003)	(0.001)	(0.171)	(0.073)	(0.140)	(0.056)
Vol	0.001	0.000	0.061	−0.151*	−0.121	−0.115*
	(0.003)	(0.001)	(0.188)	(0.080)	(0.154)	(0.062)
Capx	−0.004	−0.000	−0.348	0.010	−3.414***	−0.053
	(0.007)	(0.003)	(0.451)	(0.193)	(0.371)	(0.149)
Constant	0.058	−0.005	2.288	3.701**	10.291***	3.539***
	(0.052)	(0.023)	(3.532)	(1.515)	(2.902)	(1.164)
Sample	Family	Family	Family	Family	Family	Family
Firm effects	Yes	Yes	Yes	Yes	Yes	Yes
Year effects	Yes	Yes	Yes	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes	Yes	Yes	Yes
No. of obs	2,538	2,538	2,538	2,538	2,538	2,538
No. of firms	377	377	377	377	377	377
Prob(F)	0.01	0.01	0.01	0.01	0.01	0.01

Note(s): Table 5 presents DID estimates for any change in RPT types. DID includes the year and industry effects. Table A3 define various RPT types and firm variables. *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table 6

Effect of RPT volume with related counterparty types on the firm value in family firms

	(A) Family vs. non-family			(B) Family group vs. non-group
Transaction type	Capital assets	Operating		Capital assets	Operating
Counterparties	Holding	All	Holding	Holding	All	Holding
	(4)	(5)	(6)
Family	−0.220	−0.121	−0.116	−0.246	−0.292***	−0.159**
	(0.138)	(0.132)	(0.133)	(0.169)	(0.058)	(0.079)
CAt-2	−0.016**			−0.016*
	(0.008)			(0.008)
Family*CAt-2	0.018**			0.018**
	(0.008)			(0.008)
CAt	0.006***			0.006***
	(0.001)			(0.001)
GSt		−0.078***	−0.119***		−0.062***	−0.086***
		(0.024)	(0.034)		(0.020)	(0.026)
Family*GSt		0.082***	0.126***		0.066***	0.093***
		(0.024)	(0.034)		(0.020)	(0.026)
GSt-2		−0.002*	−0.005**		0.000	−0.005**
		(0.001)	(0.002)		(0.004)	(0.002)
Far	−0.006	0.000	0.003	−0.005	0.043	0.002
	(0.182)	(0.180)	(0.179)	(0.182)	(0.170)	(0.179)
Lev	0.261***	0.261***	0.261***	0.261***	0.267***	0.261***
	(0.073)	(0.073)	(0.073)	(0.073)	(0.074)	(0.073)
Size	−0.231***	−0.230***	−0.234***	−0.230***	−0.179***	−0.233***
	(0.043)	(0.043)	(0.043)	(0.044)	(0.041)	(0.044)
Age	−0.151	−0.095	−0.099	−0.150	−0.135	−0.097
	(0.256)	(0.257)	(0.258)	(0.256)	(0.237)	(0.258)
Vol	−0.020	−0.022	−0.023	−0.019	−0.064***	−0.021
	(0.026)	(0.025)	(0.024)	(0.025)	(0.019)	(0.024)
Capx	−0.051	−0.043	−0.054	−0.052	−0.049	−0.053
	(0.096)	(0.099)	(0.094)	(0.096)	(0.105)	(0.095)
Roa	0.884***	0.881***	0.874***	0.883***	0.858***	0.871***
	(0.201)	(0.201)	(0.200)	(0.201)	(0.212)	(0.200)
Growth	0.000	0.000	0.000	0.000	0.000	0.000
	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
Firm effects	Yes	Yes	Yes	Yes	Yes	Yes
Year effects	Yes	Yes	Yes	Yes	Yes	Yes
Industry effects	Yes	Yes	Yes	Yes	Yes	Yes
Constant	2.531***	2.276***	2.313***	2.457***	1.698**	2.278***
	(0.822)	(0.816)	(0.823)	(0.829)	(0.709)	(0.828)
No. of obs	3,468	3,467	3,467	3,468	3,936	3,467
R-squared	0.39	0.39	0.39	0.39	0.36	0.39
No. of firms	509	509	509	509	509	509
Prob. (F)	0.00	0.00	0.00	0.00	0.00	0.00

Note(s): Table 6 shows fixed effects regressions of Tobin's q (log) on RPT variables comparing family with non-family firms. Table A3 define various RPT types and firm variables. The average VIF of RPT variables is less than 2. Robust standard errors in parentheses *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Table A1

Expected direction of value effects of RPT types

RPT type	All related parties	RPT counterparty
		Holding Parties	Controlled Parties
Loans/Guarantees	Negative (H2)	Negative (H4)	Negative (H4)
Capital asset	Negative (H3)
Operating RPTs	Positive (H1)	Negative (H5a)	Positive (H5b)

Source(s): Table created by author

Table A2

Frequency of RPT types for Indian head-quartered firm

	LG	GS	CA	RPT	HCA	CCA	HGS	CGS	HLG	CLG
Total observation	4639	4662	4640	4662	4640	4640	4662	4662	4639	4639
No. of RPT >0	561	3698	1742	3855	172	1685	3250	3010	47	552
(%) of RPTs	12.10	79.32	37.54	82.70	3.71	36.31	69.71	64.56	1.01	11.90
Rs. billions	3778	7950	3039	14768	50	3014	415	7535	26.6	3755
(%) of RPT	25.60	53.90	20.60	100.00	0.34	20.4	2.80	51.00	0.20	25.43
Rs. bn firm-year	6.73	2.15	1.84	3.97	0.03	18.98	0.126	2.50	0.56	6.80

Note(s): The panel shows the frequency of RPT types for all years. The study defines variables in Table A3

Source(s): Table created by author

Table A3

Sample statistics of 578 firms

All firms				Family firms			Non-family		Non-fam vs. family
Variables	N	Mean	SD	N	Mean	SD	N	Mean	t-stat test
GS	5127	0.28	3.06	3388	0.33	3.73	1738	0.17	−1.80*
HGS	5127	0.05	1.46	3388	0.07	1.78	1738	0.02	−0.98
CGS	5127	0.22	2.40	3388	0.26	2.92	1738	0.14	−1.70*
CA	5159	0.27	6.57	3397	0.36	8.05	1761	0.09	−1.42
HCA	4639	0.07	3.58	3018	0.10	4.43	1621	0.01	−0.88
CCA	5159	0.21	3.55	3397	0.27	4.32	1761	0.09	−1.79*
LG	4616	0.01	0.05	3009	0.01	0.06	1607	0.01	−0.005***
HLG	4640	0.00	0.02	3018	0.00	0.02	1621	0.00	−0.000
CLG	4616	0.01	0.05	3009	0.01	0.06	1607	0.03	−0.005***
RPT	4805	0.12	0.45	3124	0.12	0.52	1680	0.12	−0.05
Tobq	6537	1.57	2.28	4068	1.33	1.80	2469	1.96	10.98***
Size	6829	8.29	1.66	4226	8.51	1.64	2602	7.93	−14.17***
Lev	6714	0.29	0.38	4181	0.35	0.41	2533	0.19	−17.13***
Roa	6800	0.14	0.13	4203	0.12	0.13	2596	0.16	9.99***
Far	6790	0.30	0.20	4197	0.32	0.19	2592	0.26	−11.76***
Sta	6815	1.36	21.35	4222	1.48	26.72	2592	1.17	−0.57
Vol	6446	0.37	1.11	3995	0.35	1.04	2450	0.39	0.040
Capx	6245	0.05	0.31	4077	0.05	0.16	2168	0.04	−0.014*

Note(s): The variables are: GS: buy/sell of goods and services, i.e. operating RPTs over sales; HGS: operating RPTs with holding parties over sales; CGS: operating RPTs with controlled parties over sales; CA: capital asset RPTs over sales; HCA: capital asset RPTs with holding parties over sales; CCA: capital asset RPTs with controlled parties over sales; LG: outstanding RPLGs over assets; HLG: RPLGs to holding parties; CLG: RPLGs to controlled parties; RPTs: (log) total RPTs over assets; Tobq: (log) ratio of the sum of market cap and the book value of debt over assets; Size: log of assets; Lev: total borrowing over assets; Roa: EBITDA over assets; Far: fixed assets over total assets; Sta: sales over total assets. *** p < 0.01, ** p < 0.05, * p < 0.10

Source(s): Table created by author

Notes

1. Beginning with Satyam, RPTs are behind several recent scams in India. Gitanjali Gems, Winsome Diamonds, Sterling Biotech, DHFL and many more siphoned off billions of rupees through related parties. USL Ltd., the largest spirit maker, gave ₹ 12.25 billion in unsecured loans to related firms to be never repaid. Jet Airways, despite losses, gave ₹ 18 billion as loans to JetLite.

2. I use these criteria based on past criminology research that has identified six attributes that enhance the chance of theft of a product. The product should be concealable, removable, valuable, available, enjoyable and disposable (Clarke, 1999).

3. Anecdotal evidence from speaking to several executives/auditors suggests that phantom yet small transactions about rental, consulting fees, reimbursement of perks by directors/owners/relatives are quite common in Indian family firms.

4. I avoid the log form of LG as it drops a large majority of observations and makes the panel data analysis weak.

5. I use the log form since capital RPTs are dispersed and very large in some firm years. The log form provides the best fit and reduces multicollinearity between current and lagged terms.

6. Satyam, listed on the NYSE, was the fourth largest Indian software firm with operations worldwide. It was a family-controlled firm founded by its chairman Ramalinga Raju. However, it had stellar independent directors such as Krishna Palepu from Harvard and Vinod Dham - the father of Pentium. Four months ago, The World Council for Corporate Governance had awarded Satyam with the Golden Peacock Global Award for Excellence in Corporate Governance. Investor Relations Global Rankings had rated it with Best Corporate Governance Practices for 2006 and 2007. However, its stellar board on December 16, 2008, approved the $1.2 billion cash acquisition of two family firms. The RPT was called off after protest from institutional investors. On January 7, 2009, Ramalinga Raju disclosed that the firm had been fudging its accounts for years. Its cash was mostly non-existent.