Let’s cut to the chase. I’m a big fan Darrel Duffie’s work on the mechanics of markets. It’s something that hasn’t featured prominently during my Financial Economics masters, but it’s really worthwhile to think about. Not all assets trade on highly liquid and near frictionless exchanges. In fact, most assets trade on (intermediated) OTC markets which has implications for price-discovery and efficiency. It’s not surprising that a large strand of the literature has devoted time and effort researching how the design of a market or intermediation by financial institutions influences asset pricing behaviour of particular asset classes (more on such anomalies in a later blog post!). Here comes a brief state-of-the art introduction on OTC markets.
What are OTC markets?
To illustrate how markets mechanics can affect asset prices, we take three types of markets as a starting point, 1) a centralized market and 2) an over- the-counter market (OTC market) and 3) a hybrid OTC market. For now these markets will suffice (but of course more complex types of markets exists).
Figure 1 describes a centralized market. The red dot is the exchange which keeps the order book. As such, this central marketplace intermediates transactions in the market. One can think of for example of the New York Stock Exchange (NYSE). The blue dots are buyers or sellers of assets traded on the exchange and they engage in trading by e.g. sending limit or market orders to the order book.
Not all trading occurs on exchanges. Figure 2 describes an OTC market. As opposed to a centralized market, transactions on an OTC market are not intermediated by a central marketplace and as a result an OTC market exhibits a natural friction: in order to engage in transactions, the blue dots have to shop around i.e. it takes some time to find a counterparty (this process can take longer than one might expect!).
To ameliorate some of this friction there exists hybrid forms such as shown in figure 3. There the red dots represent e.g. dealers in the form of investment banks, such as Morgan Stanley or J. P. Morgan. These dots intermediate transaction between buyers and sellers. These hybrid forms such as in figure 3 are often a better description of most OTC markets in real life.
Relevance of OTC market efficiency
A key advantage of OTC markets however is that market participants are free to negotiate any mutually attractive deal. That’s partly the reason why OTC markets are so big. Far bigger than exchange traded markets. To give some sense of magnitudes: almost all of the $92t outstanding in bonds is handled by OTC markets and essentially all trades in currencies, commodities, repos, securities lending, bank loans and private equity are handled by OTC markets (Nagel, 2016)
OTC markets tend to be concentrated as well. A small number of dealer banks intermediate almost all OTC swaps (Nagel, 2016) and the majority of OTC trades in repos, securities lending, FX forwards, equities, currencies (Rime and Schrimpf, 2014) and corporate bonds (Biais & Green, 2007).
Impediments to OTC market efficiency
The literature distinguishes broadly four sources of limitations to OTC market efficiency. They sometimes overlap but nevertheless distinguishing among them is still useful.
Sequential trade of assets by search and bargaining
As mentioned earlier imperfect search friction are natural to dealer-intermediated and OTC markets. Duffie, Garleanu and Pedersen (2002, 2005, 2007) basically describe two problems arising from the search friction. First they describe a hold up problem in OTC markets. Meaning that trades that should have occurred won’t have occured in the market: some investors will not find the terms of trade good enough to enter the market in the first place. The second issue arises from the search costs. Typically in OTC markets investors will face a trade-off between execution cost and speed. And indeed, in a recently published empirical paper in the JoF, Li & Schürhoff (2018) found evidence for such trade-off to exist in the US OTC municipal bond market.
Another source of friction is opaqueness. Opaqueness in essence is a violation of the Law of One Price (Green, Hollifield & Schürhoff, 2006). For some two asymmetrically informed investors about the value of the asset, and because of the particular structure of the market, assets trade at radically different prices almost simultaneously. Opaqueness in the market is indicated by price dispersion: the range of prices of the same asset, traded at a given point in time, between different pairs of counterparties of the market.
A key feature opaqueness is that often there is no ongoing price in the market. Which again leads to hold up problems as mentioned earlier. That hold up problem can be mitigated by post-trade price transparency as shown in Asquith, Covert & Pathak (2013). In their paper they show empirically that regulation enforcing post-trade price transparency, affects bid-ask spread and volume in the corporate bond market. Another solution shown in Duffie, Dworczak & Zhu (2017) is to implement a benchmark. They argue that announcing the ongoing price in an intra dealer market benefits entry. A particular investor would be more likely to enter the market if the price was amenable to him. In part because now the investor knows what the ongoing price is and in part because when the investor faces a counterparty, the counterparty knows that the investor knows the ongoing price. The counterparty can’t exploit the investor, which leads to the counterparty narrowing his bid-ask spread. This makes it more likely for an outside investor to enter.
Another argument Duffie, Dworczuk & Zhu (2017) provide is that benchmarking would lead to better matching. If the counterparty will not give the investor a good price then the investor is going to guess that the counterparty has a high cost for that particular asset at that particular time, and so the investor will shop around. And vice-versa if the counterparty gives him a good price the investor will not shop around.
Limits to dealer formation and exchange migration
This strand of the literature tries to explain how OTC markets emerge and what impedes certain OTC markets to transitioning to low-cost centralized trading. E.g. Lee & Wang (2018) that OTC markets allow OTC dealers to price-discriminate among informed and uninformed investors. Dealers earn a a large fraction of the volume of trade they intermediate. And they earn large bid-ask spreads on that volume of trade. If the market transitions to centralized trading, they would lose a large fraction of the trades to direct customer to customer trades and their bid-ask spread would be smaller. So its a lose lose development for the dealers to encourage the development of exchange traded markets.
High cost of dealer balance-sheet “space”
Andersen, Duffie & Song (2018) is one of the few theoretical models on how capitalization of dealers is an important determinant of risk premiums in markets (also true for exchange traded markets). The empirical literature corroborate these findings (but I won’t delve into that in this blog post).
For illustration purposes, we take the case of covered interest rate parity violations. The LHS of figure 4 show the difference in cost of borrowing directly in US dollars funding markets versus the cost of borrowing indirectly in a foreign currency and swapping it back to US dollars using the derivatives market. If markets were working perfectly the difference should be zero. But they don’t. This is called the cross-currency basis.
Notice too that pre-crisis the cross-currency basis was tiny, corresponding to low credit spreads for dealers (the spreads were small because the dealers were too big to fail). Because the balance sheet cost was so small, dealers would absorb any balance sheet expanding trade that would earn the dealer just a few basis points in order to capture this arbitrage. Post-crisis the cost for balance sheet space (in terms of cost for funding) increased enormously (see RHS of figure 4) and as a result the cross-currency basis widenend.
In Anderson, Duffie & Song (2018), it is the credit spread on the RHS of the balance sheet that are causing the big cost of balance sheet space post-crisis (and not regulatory capital requirements).
Suppose dealer A is considering the following trade. On the RHS of figure 5 the dealer borrowed directly in US dollars. With those funds the dealer purchases EUR assets and swaps it back to US dollars on the derivative market (seen on the asset side of the balance sheet). Because the dealer added a relatively safe asset to its balance sheet, he has improved the position of his legacy creditors (see liability side of the balance sheet). These creditors like this trade and as a result the legacy debt increases in market value (dark blue area increases in size).
If there were no arbitrage profit (meaning if the asset traded at fair value), the equity owners just lost money (notice the area between the upper horizontal line and green rectangle in figure 6, this is called the Funding Value Adjustment in the industry). The dealer is only going to execute this trade if the arbitrage profit is larger than the FVA i.e. if the dealer can purchase the asset at below market value.
If the dealer is not well-capitalized, the dealer will be discouraged to offer narrow bid-ask spreads. So the cost of dealer balance sheet space is another source of friction in OTC markets.
In comparison with research on efficiency of exchange traded markets, research on OTC-markets efficiency is relatively recent. This is partly due to the fact that gathering data on OTC-markets is by definition more difficult than exchange traded markets. However given the size and concentration observed in OTC-markets, it’s worthwhile to pursue more research on financial intermediation in OTC-markets.
Andersen, L., Duffie, D., & Song, Y. (2019). Funding value adjustments. The Journal of Finance, 74(1), 145–192.
Asquith, Paul, et al. “The market for borrowing corporate bonds.” Journal of Financial Economics 107.1 (2013): 155–182.
Biais, B., & Green, R. C. (2007). The microstructure of the bond market in the 20th century. Tepper School of Business, 134.
Du, W., Tepper, A., & Verdelhan, A. (2018). Deviations from covered interest rate parity. The Journal of Finance, 73(3), 915–957.
Duffie, D., Dworczak, P., & Zhu, H. (2017). Benchmarks in search markets. The Journal of Finance, 72(5), 1983–2044.
Duffie, D., Garleanu, N., & Pedersen, L. H. (2002). Securities lending, shorting, and pricing. Journal of Financial Economics, 66(2–3), 307–339.
Duffie, D., Gârleanu, N., & Pedersen, L. H. (2005). Over‐the‐counter markets. Econometrica, 73(6), 1815–1847.
Duffie, D., Gârleanu, N., & Pedersen, L. H. (2007). Valuation in over-the-counter markets. The Review of Financial Studies, 20(6), 1865–1900.
Green, R. C., Hollifield, B., & Schürhoff, N. (2006). Financial intermediation and the costs of trading in an opaque market. The Review of Financial Studies, 20(2), 275–314.
Lee, T., & Wang, C. (2018). Why Trade Over-the-Counter? When Investors Want Price Discrimination. Working Paper.
Li, D. and Schürhoff, N. (2019), Dealer Networks. The Journal of Finance, 74: 91–144. doi:10.1111/jofi.12728
Nagel, J. (2016). Markets Committee Electronic trading in fixed income markets. Tech. Rep. 9789291974207, Bank for International Settlements.
Rime, D., & Schrimpf, A. (2013). The anatomy of the global FX market through the lens of the 2013 Triennial Survey. BIS Quarterly Review, December.