Imagine an asset class with a decently positive expected rate of return, little to no equity beta, and little to no interest rate duration. A unicorn? We think not.
It’s no secret that we at GMO have concerns today about the valuations of traditional asset classes. The S&P 500, as measured by the Cyclically Adjusted Price to Earnings Ratio, sits above its 2007 levels and within a whisper of levels seen just before the crash of 1929. Our published forecast for US large cap stocks, and many other parts of the global equity markets, sits in solid negative territory.
For bonds, US 10-year treasury yields hit 1.36% only 12 short months ago, the lowest ever recorded in American history. And last spring, interest rates in Europe were the lowest in four centuries. (Dare I even mention that the 50-year Swiss bond’s nominal yield went negative?) You cannot make this stuff up. Our forecasts for most bond markets around the planet are pretty bad, as any form of normalization of rates is likely to cause some pain.
When we see both equities, generally, and bonds, generally, with poor to even negative expected real rates of return, we ask ourselves a simple question: Is there an asset class out there that has the following attributes:
a) a decently positive expected rate of return;
b) little to no equity beta; and
c) little to no interest rate duration?1
For a significant chunk of the last 150 years of US capital markets history, the answer was typically… cash! (Please note that in 2007, when we were warning about a global risk bubble and negative expected rates of return for almost all risk assets, T-Bills were yielding 4.7% nominal, or almost 2% real!) Today, we cannot say that. T-Bills, even with the recent Fed hikes, are still paying a negative real yield.
So, we ask again: Is there another asset class out there with these seemingly magical attributes? Or are we searching for a unicorn?
No unicorns here. No magic required. It’s called Merger Arbitrage (Merger Arb, for short).
Positive expected return
First, let’s do a quick review of what a Merger Arb portfolio typically holds. We’ve provided a pretty clean example in Exhibit 1. Imagine that company “A” (Acquirer) makes an offer to buy a publicly traded company “T” (Target). Let’s assume it’s an all-cash deal. Exhibit 1 lays out the sequence of events. In Step 1, Company T is trading for $75 before the announcement. In Step 2, Company A announces to the world that it wants to buy Company T for $100, or a premium of $25. Within seconds, the price of T’s stock moves upwards; it gets very close to the $100 offer price. Does it go all the way? Typically not. In our example, the price moves to $93. Why? Because the deal typically takes six to nine months to actually close and during that period some bad things could happen. For example, a regulatory body might put a stop to the deal; the Acquirer might not be able to line up the financing for an all-cash deal; or due diligence unearths some hidden liabilities connected to Company T. You get the idea – there is some uncertainty, or risk, that the deal collapses, or “breaks,” in industry parlance. This $7 spread is what Merger Arb players are interested in: They would step in to buy Company T for $93, hoping to earn the $7 spread as compensation for taking on those latent risks. This, of course, is the arbitrage.
Over the next few months, the price of T will oscillate, depending upon the market’s changing perceived odds that the deal will close or break. The return profiles of each outcome are strikingly different. In Step 3a, we suppose the deal closes: T’s price moves to $100, and the Merger Arb investor books a gain of 7.5% (a $7 return on an initial $93 investment). However, if the deal breaks (Step 3b), T’s price will likely collapse to the pre-deal price of $75, costing the investor $18, or -19.4%. Now, why would any investor do Merger Arb given the asymmetry of a modest gain relative to the risk of a sizable loss? The answer lies in what we call “Expected Return.” As it turns out, Merger Arb deals close more often than not historically, with roughly 90% of announced deals actually closing.2 In our example, we can now calculate a simple Expected Return, given these probabilities. (See the blue box in Exhibit 1, where we multiply the probabilities of winning or losing times their respective gains or losses, or +4.8%. Not bad.)