سود ناویژه در بازار مشتقات و ثبات بخش بانکی

We investigate the effects of margining, a widely-used mechanism for attaching collateral to derivatives contracts, on derivatives trading volume, default risk, and on the welfare in the banking sector. First, we develop a stylized banking sector equilibrium model to develop some basic intuition of the effects of margining. We find that a margin requirement can be privately and socially sub-optimal. Subsequently, we extend this model into a dynamic simulation model that captures some of the essential characteristics of over-the-counter derivatives markets. Contrarily to the common belief that margining always reduces default risk, we find that there exist situations in which margining increases default risk, reduces aggregate derivatives’ trading volume, and has an ambiguous effect on welfare in the banking sector. The negative effects of margining are exacerbated during periods of market stress when margin rates are high and collateral is scarce. We also find that central counterparties only lift some of the inefficiencies caused by margining.

For a long time, margining, a mechanism for attaching collateral to derivatives contracts, was considered a panacea to mitigate default risk in derivatives markets (International Swaps and Derivatives Association, 2005). However, in many financial calamities during the past few decades, including the collapses of Metallgesellschaft, Long-Term Capital Management (LTCM), and more recently Bear Stearns, Lehman Brothers, and American International Group (AIG), margining played an ambivalent role. It is not unreasonable to believe that it exacerbated the recent financial crisis. At present, it is an open question as to what the overall effect of margining is in a financial system and in an economy more generally.1 In this article, we identify situations in which margining of derivatives, two-way contracts in which both parties are both potential creditors and potential debtors, decreases trading volume, increases default rates and default severity, and reduces welfare in the banking sector. Our analysis shows that margining presents derivatives counterparties and regulators with a delicate trade-off. On the one hand, margins reduce default severity by reducing banks’ exposure to the default of their counterparties. On the other hand, margin requirements generate several types of costs, particularly when banks use derivatives for hedging purposes. First, by imposing a funding constraint on banks’ trading strategies, margin requirements can limit the number of derivatives contracts traded by a bank and thus can prevent it from implementing its optimal hedging position. Second, increased margin requirements can indirectly constrain a banks hedging strategy by reducing the number of contracts outstanding of other banks. Moreover, increased margin requirements can reduce the credit quality of a bank’s counterparties (that is, increased probability of default and loss-given-default) by constraining the counterparties’ hedging strategy. These considerations lead to the conclusion that margining affects market outcomes through several different direct and indirect channels that may interact in subtle ways, some of them counter-intuitive. Moreover, it may impose negative externalities, that is, negative, indirect effects on other parties within the financial system or the economy, that are not transmitted through prices. In the remainder of this article, we address the following research question: How do the various margining mechanisms observed in current derivatives markets affect trading volume, default risk, and welfare in the banking sector, in particular, during periods of market stress? To analyze the effectiveness of margining within the banking sector we present two models. Our first, baseline model is a simple, static equilibrium model in which we develop some basic intuition for the main channels of margining. Subsequently, we present a dynamic market model that extends the equilibrium model. The dynamic model captures many features of modern derivatives markets so as to analyse the various channels of margining and their interaction in a more realistic setting. However, this model cannot be solved analytically; hence, we evaluate it using simulations. In the equilibrium analysis, our baseline model, we consider a one-period economy with an incomplete market and two (groups of) risk-averse banks. The banks have opposite endowments in a long-term, illiquid asset and a certain amount of cash. They wish to hedge the risk of their endowment by trading short-term derivatives contracts with each other. Banks maximize the expected utility of wealth by choosing the optimal number of derivatives contracts. Because markets are incomplete, we allow the banks to default. We then introduce a margin requirement aimed at mitigating default risk associated with the trading of the derivatives contracts. Solving for the banks’ optimal trading strategies, we analyze the impact of margin requirements on their welfare (as measured by their utility of wealth), their default risk, and on the volume of derivatives traded. We find that exogenously imposed margin requirements can be privately and socially sub-optimal. Indeed, using numerical analysis, we find that sometimes a margin requirement of zero is optimal. The negative effects of margining increase as the constraints imposed by the margin requirements on the banks’ optimization problem tighten. More important, our results also suggest that when banks differ in key characteristics that affect the optimal level of the margin requirement, including their probability of default and risk aversion, privately and socially optimal levels of margining may not be the same, which in turn implies that the level of margining in a market will affect not only aggregate welfare but also the relative distribution of welfare. Subsequently, we extend our baseline model to create a more realistic simulation model of derivatives trading in the banking sector. More precisely, we analyze a market consisting of several heterogeneous banks that face a similar optimization problem as before while assuming that the banking sector is experiencing severely adverse market and credit risk conditions. The latter assumption is made in order to determine how margin requirements affect this banking economy during market crises which are often deemed to represent the market conditions during which collateral is most valuable. In order to make the model more realistic, we calibrate its parameters with actual derivatives market data. We use this model to study the effects of initial margin, variation margin, and a central counterparty on market outcomes. We find that the introduction of margining, both in the form of initial and variation margin, significantly deteriorates derivatives market liquidity while it increases banks’ default rates and ambiguously affects their welfare when assuming a mean–variance utility function. These results are more pronounced when initial margin levels are strengthened. The simulation results thus support the results obtained with our baseline model regarding the impact of margining on banks’ welfare and derivatives trading liquidity. They further show that, under stress scenarios, tighter margin requirements will even exacerbate banks’ default risk. In all our analyses, initial margin levels are set ex ante and remain constant; that is, we exclude pro-cyclical adverse effects of margining due to increases in margin rates during periods of stress. Our results are reminiscent of the theory of ‘second best’, according to which the elimination of a market imperfection does not necessarily make an economy better off in the sense that it can exacerbate the negative effects of other market imperfections. We believe that our results explain some of the existing empirical research in this field. Hartzmark, 1986 and Hardouvelis and Kim, 1995 found that increases in margin rates at major derivatives exchanges led to a decrease in trading volume and open interest. We also believe that our results explain, at least in part, current events in financial markets such as the collapse of AIG.2 Hence, our results for both the baseline equilibrium and the extended simulation models presented in this article are of interest to public policy makers, especially in light of the recently increased use of margining in over-the-counter (OTC) derivatives markets and its ambiguous effects on welfare in the banking sector. Indeed, we find that almost perfect coverage of counterparty default risk exposure by margining is sub-optimal during periods of market stress. This finding is relevant to the role of derivatives trading, including credit derivatives trading, in the recent liquidity and credit crisis in the global banking sector.3 Our results also emphasize the significance of the interdependence between different types of risk, such as credit and liquidity risk, suggesting that these risks should ideally be analyzed and managed jointly rather than separately. Therefore, any change in margining policies in financial markets, such as the introduction of a central counterparty, should be considered carefully. At the same time, margining should become a key issue in the design and implementation of financial market policies, as suggested in Turner (2009). We proceed as follows. In Section 2, we briefly review the literature relevant to this study. In Section 3, we develop our baseline model, a simple equilibrium model to shape the intuition for the various channels of margining. In Section 4, we extend this baseline model to create a more realistic, dynamic simulation model. Finally, Section 5 concludes the study.

In this study, we examined the impact of margin requirements on derivatives contracts and on banks’ welfare, trading volume, and default risk. Both the baseline and the extended simulation models presented in this study suggest that margin requirements may have a negative impact on banks’ welfare, increase their default risk and lower their trading volume in derivatives contracts. These adverse effects are particularly acute when margin rates are high and collateral is scarce. The extended model further illustrates that these effects are most likely to prevail during periods characterized by highly volatile interest rate and credit markets. In some sense, an increase in margin requirements is then comparable to a bank run in that a bank is forced to deliver cash to its creditors, cash that it does not have at hand. The relation between margin requirements on one side and trading volume, default rates, losses-given-default, and banks’ welfare on the other can, in principle, be tested empirically. Hartzmark, 1986 and Hardouvelis and Kim, 1995 and others investigated the relationship between margin requirements and open interest as well as trading volume for several derivatives exchanges. They found that an increase in margin requirements tends to have adverse effects on both open interest and trading volume. The theoretical model and the simulation results presented in this study offer a potential explanation of these findings. To the best of our knowledge, though, no previous empirical analysis has so far examined the relation between margin requirements and default risk as well as its various constituents. It should be clear from the analyses presented that the effectiveness of margining should be evaluated not only in terms of its effects on credit exposure but also in terms of its impact on banks’ welfare. Of particular interest is the search for default risk mitigation mechanisms that preserve the welfare benefits of margining while reducing its costs. One such mechanism is rehypothecation, which allows market participants to re-use the collateral they receive to serve margin requirements. Rehypothecation can significantly reduce the amount of collateral necessary to cover a given portfolio of contracts. Rehypothecation has two major drawbacks, however. First, it does not completely eliminate the credit exposure of a given position, as described in Section 3. Second, it exposes the collateral posted to default risk, as some hedge funds painfully experienced in the case of Lehman’s collapse.20 Another mechanism that can be used to alleviate the burden of margining is the introduction of a central counterparty. A central counterparty provides the same benefits as rehypothecation. It typically reduces credit exposure even further and can at the same time further decrease the margin requirement of a given position. In the context of the model presented in this study, the advantages of a central counterparty manifest themselves mainly through a reduction of the average losses given default; the impact on aggregate default risk remains marginal. Central counterparties, however, provide additional benefits. Most important, they provide facilities, such as default funds, to mutualize losses in relation to counterparty defaults, thereby reducing moral hazard. In addition, by centrally processing transactions, they reduce legal and operational risk. Another important advantage of a central counterparty over other default risk mitigation mechanisms is the centralization of information about market participants’ positions that the central counterparty can channel. This in turn permits centralization of aggregate default risk management, which can be desirable from a macro-economic policy perspective. In OTC markets, the distribution of risk is typically unknown, rendering default risk management cumbersome, particularly in periods of market stress. Schinasi (2006) claimed that these information asymmetries in OTC markets constitute a major threat to the stability of the financial system. Central counterparties thus offer benefits that can be significant in that they can effectively reduce some of these informational asymmetries and potential related externalities. As discussed earlier, central counterparties have the potential to improve transparency of trading positions in a market, which in turn would enable market participants to take this information into account when valuing contracts, thereby internalizing potential externalities (Acharya and Bisin, 2011). Let us conclude with two final remarks based on this study’s findings. First, the choice and the implementation of alternative default risk mitigation mechanisms in derivatives markets can have a significant impact on the welfare and the risk profiles of banks trading in these markets. Second, in this study, margin rates were set exogenously. This observation leads to the following questions: (1) What is the optimal level of margining for an economy? and (2) How does welfare change if margin rates are determined endogenously as part of the banking system’s optimization problem? We hope that these important questions will be addressed in further research.