Copyright 2014

What Has Become Of Risk Arbitrage?

Some years ago I wrote a book on investment strategy, which included this deathless prose:

Where the values of two assets are related to each otherhellip;.[a]rbitrage involves going long the asset that is valued too cheaply and selling short the asset that is too dear, and thus is reliant on a perception of where their values should be relative to each other.

But as always, the devil is in the details: hellip;two assets are related to each otherhellip; and hellip;where their values should behellip; are both pretty glib. Of course, some arbitrage is straightforward - for instance, between gold on different trading venues. These are perfect arbitrages, where there is a systematic reason why the value relationship between the two of the trade should be stable, so that if their prices depart from equilibrium, there is a riskless profit to be made.[i] As with any riskless opportunity, these arbitrages tend to be of short duration and offer narrow profit margins. They only attract those who can leverage them extravagantly and can transact nearly cost-free.

There are plenty of imperfect arbitrages. I will concentrate on risk or merger arbitrage here, but there are a host of different problems currently affecting other arbitrage strategies, too (see here). Risk arbitrage exploits the relationship between the securities of an acquirer and its target - primarily, but not always exclusively their equities - which is established by the terms of the acquirers bid. Obviously, the risk to this strategy is that the price relationship that it seeks to exploit can break down - for instance because the target successfully defends itself, because the acquirer cannot obtain financing or because of regulatory interference with the deal. Hence these trades are far from riskless, and are correspondingly less leveraged than perfect arbitrages.

Academic finance discussions often treat arbitrage as a sort of natural magic that saves theory from inconvenient data. But in fact, arbitrage is a business, not so terribly different from any other: it is constrained by the number of opportunities available for it to exploit, by the potential margin on those opportunities (their richness), by operational items such as transaction costs, the cost and availability of credit and security lending as well as the cost of the research required to identify opportunities. Arbitrage will not occur if the frequency or richness of the opportunities to engage in it are insufficient to pay for the cost of identifying and exploiting them. In risk arbitrage, research costs include proprietary securities analysis and, almost inevitably, legal advice regarding a range of issues from aspects of the targets Articles of Incorporation to applicable regulation.

Opportunities for risk arbitrage are fairly plentiful at present:

(click to enlarge)

2015 will not be a record year in terms of the number or value of deals, but we are by no means amidst a deal-making drought. Dealogic reports that, at $967 million, the average size of transactions is the second highest ever. Prospects for 2016 look fairly sound, since little has occurred over the course of 2015 to change the background conditions for Mamp;A. A bit less than half of the value of Mamp;A activity through October this year has been in North America.

The richness of these opportunities is another matter. There are several large - and expensive - databases on deal statistics, but they are not in fact very enlightening. Control premia vary all over the map, by company and industry, with time and by metric employed (bid value relative to share price, sales, earnings, assets, etc.). They tend to be lowest at the trough of a recession, when bargain-hunters can snap up companies in distress at little or no premium. They tend to be highest when private equity firms lose discipline and bid indiscriminately. Although, as I have written elsewhere in SA, private equity coffers are very full at present, the desperation to deploy those assets that we have seen in previous cycles has not materialized. We are currently in neither situation. Average share valuations are relatively high, creating an additional discouragement to generosity, if one were needed. So most deals being proposed at present offer adequate opportunities to risk arbs. With the number and size of deals healthy, the premia attached to them reasonable (although apparently declining), the general conditions for revenue capture in the risk arbitrage business are good.

As in any other commercial activity, the size of transactions and the gross margins achievable on them do not, in themselves, determine success: arbitrage is a competitive business, and the competition seems to have been increasing. Unlike, say, long-only investing in equities, but similar to holding bonds to maturity, arbitrage is what I call a finite trade: the maximum potential profit - the spread - is known when the trade is initiated. So, naturally, as more traders seek to exploit a deal, the premium available to each of them is diluted.

It is difficult to determine how many risk arbitrageurs there are, and how much capital they have at their disposal. This is partly because, during times when deals proliferate and spreads are wide, a considerable amount of capital flows into the trade. Although this also reflects the re-opening of existing funds and the launch of new ones, much of it is due to the reallocation of capital by event-driven managers, who may engage in a wide range of trades, among them merger arbitrage. Until 2014, many were concentrating their efforts and capital on various permutations of distressed investing. Since the Crash created numerous distressed investing opportunities, which these firms were noticeably effective in exploiting, event-driven hedge funds attracted a great deal of capital. However, these trades have become fewer and less productive, and these firms capital has gradually been deployed in risk arbitrage.

Investment banks participation in risk arbitrage has declined steadily. Much of this has to do with the dislike their Mamp;A bankers clients expressed for the activities of their trading desks (which, of course, included shorting their shares), but the regulatory onslaught against banks principal activities has not helped, either. Virtually any risk arbitrage unit within a bank that found itself de-funded could always hive itself off as a fund, and most of them have. They probably no longer have access to anywhere near the capital they had while affiliated with a bank, and their transaction costs are no doubt higher, but they are no longer subject to the compliance and conflict of interest considerations that would have hampered their trading while employed by a bank.

With the decline of banks arbitrage units, there are no longer any traders in the merger market with almost unlimited access to low cost leverage. Since the Crash, the availability and cost of leverage has been a significant constraint on the activity of most hedge funds. While risk arbs are not typically as leveraged as their brethren in many relative value trades, four to five times is probably what most of them want, if they can get it. This is essential: spreads are rarely so wide that they would be worth the trouble and risk involved in capturing them if the trade could not be leveraged. The first trader or two to respond to a merger announcement might be able to lock in good returns without leverage: all the rest must rely on it to achieve their required returns.

Risk arbitrageurs are not the only businessmen to find it harder to obtain credit - in their case, from prime brokers - since the Crash. While this may tend to reduce competition among traders (including by driving some out of the strategy entirely), it also causes them to select their targets more carefully. One such consideration is the ease and cost of borrowing the shares that constitute the short leg of the trade - a separate issue from the availability of leverage. Since the Credit Crunch, there has been a bewildering increase in securities lending regulation worldwide. Many suppliers of lending inventory have re-examined their practices, and some have decided, like Vanguard, to restrict their lending. Others have stopped lending completely. Some intermediaries have also withdrawn from the business. Sufficiently widespread adoption of such strategies tends, of course, to raise the cost of all security loans. Finally, increased competition for stock loans tends to raise costs and reduce availability.

Securities are borrowed for many reasons, and risk arbitrage is only one of them; however, over periods longer than about a week (which washes out market makers transitory borrowing activity), changes in loans outstanding are primarily influenced by hedgers and arbitrageurs borrowings.

(click to enlarge)

At first glance it seems that the situation is quite healthy: equities available to lend increased by 13% and stocks out on loan (at least partly reflecting rising Mamp;A activity) by 36%. But since stock prices rose some 15.5% during the period, the amount of shares available to borrow actually declined by about 2.5% during the period shown. Note also that securities out on loan rose from 26% to 31% of those available to borrow: this suggests a fairly tight market.

Arbitrageurs legal costs have risen, too. With ever more erratic Justice Dept. application of anti-trust statutes, with other government offices claiming the right to review transactions, and new legal theories fostering ever more imaginative construction of laws whose import was once well understood, demand for legal research is a growing. Regulatory risk has always influenced arbitrageurs choice of trades, but these risks have become unpredictable. In large deals which involve many jurisdictions, such as the AB Inbev (NYSE:BUD)/SABMiller (OTCPK:OTCPK:SBMRY) transaction, they may have become imponderable. Is there such a thing as Chinese anti-trust policy, and can anyone guess how it might be implemented? This uncertainty is reflected in the failure of spreads on many transactions to narrow over time, including deals with little or no transnational dimension, such as those between Office Depot (NASDAQ:ODP) and Staples (NASDAQ:SPLS), Anthem (NYSE:ANTM) and Cigna (NYSE:CI), Walgreens Boots (NASDAQ:WBA) and Rite Aid (NYSE:RAD) and finally Baker Hughes (NYSE:BHI) and Halliburton (NYSE:HAL). Persistently wide spreads suggest skepticism as to whether a deal will be consummated. US authorities seem to aspire to Chinese levels of opacity.

Of course, there may be an opportunity to jump into a deal later. Arbitrageurs do not just build positions when a transaction is announced and hold them until it closes. While most of them retain such a core exposure to a transaction, they trade actively around that position. Between offer and closing there are usually developments that make closing look more or less likely, and the spread widens or contracts in response, offering trading opportunities that most arbitrageurs exploit. The aggregate profits on these trades may in fact exceed the gain achieved on the core position. Naturally, this circumstance reinforces risk arbitrageurs preference for deals between companies whose shares can be traded actively and borrowed easily.

As a result, we have seen various transactions, including the AB Inbev/SABMiller deal mentioned above, and Weyerhaeusers (NYSE:WY) bid for Plum Creek (NYSE:PCL) where arbitrageurs seem to have lost control of the situation, if they ever had any to begin with. Although a new bidder is almost impossible in both situations, the stocks are trading as though one is expected. This is because arbitrageurs presence in these transactions is relatively small, although for quite different reasons.

(click to enlarge)

There are a number of ways that investors who do not have access to hedge funds or balk at their fees can obtain exposure to merger arbitrage. To my mind, none of them is very satisfactory. SA contributor Sam Kovacs has pretty definitively made the case against the two ETFs available, the IQ Arb Merger Arbitrage ETF (NYSEARCA:MNA) - which does not in fact engage in arbitrage at all - and the ProShares Merger ETF (BATS:MRGR) - which does, but according to an algorithm that is unlikely to produce acceptable returns, especially under current merger market circumstances.

Among conventional mutual funds, there are the Touchstone Merger Arbitrage Fund (TMGAX), the Merger Fund (MERFX) and the Arbitrage Fund (ARBFX). These funds do what they say on the box, and have fairly consistent performance, but because they are not significantly leveraged, it is consistently in the low single digits. Those rates of return do not strike me as a sufficient to compensate for the risk of drawdowns greater than their average annual returns, which all three have experienced at least once. I have frequently argued that hedge fund indices have serious flaws, for instance here, and that they are not good for much other than benchmarking returns for a single time period. But these funds returns do not compare favorably even with such an imperfect metric (note that the time period illustrated is the entire period for which daily data on the index are available: both MERFX and ARBFX have been in existence longer, while TMGAX was launched in 2011):

Various other funds, which are not purely risk arbitrageurs but include that strategy among others that they pursue, include Gabelli ABC Fund (GABCX), Quaker Event Arbitrage (QEAAX) and AQR Diversified Arbitrage Fund (ADANX). Of these, only GABCX has produced returns and volatility that approach those of the index illustrated above.

[i] There is no such thing as a riskless trade. These arbitrages earn that description because there is essentially no risk that the trades return driver - the correlation between the two assets - will fail. In fact these trades carry significant risk, but it is operational rather than economic.