A good proportion of the portfolio is dedicated to "arbitrage". Now, in the last few decades almost all relative value trading has been called arbitrage of some kind, ie. buying very 'cheap' value shares is now "time arbitrage". I use the term slightly more tightly than that, but more loosely than buying and selling exactly the same instrument, at exactly the same time for a penny difference in profit.
When I say arb, I mean things like merger arbitrage, credit basis trades, rights issues, convertible bond arb (with everything hedged!). These strategies are first order market neutral, ie. their pnl doesn't show any correlation to small moves in the overall market. They do tend to show correlation over big moves to the downside (which is what you really care about).
Some strategies have a fundamental reason for the weak downside beta. For example the probability that an acquirer walks away from a deal increases when the economy changes for the negative, the credit basis can often move in unexpected ways because of differing funding spreads and expectations of restructuring, etc.
In general there is also a second mechanism, liquidity and contagion. Holders of these arb, basis, RV trades tend to have similar positions and when the market takes a real beating there is often an increase in risk premiums and blow out in relative value spreads (whether fundamentally justified or not) as investors look to increase their liquidity. Additionally you can have real contagion as any stat arb fund that was around in 2007 can tell you, ie. a fund that runs several strategies blows up in one more directional one and then has to liquidate all their strategies, that can set off a cascade of liquidation.
The general point that came up in conversation however is that with all these type of arb spreads, one is short a put on liquidity. That diversification is an illusion, and if anything its better to panic first.
No comments:
Post a Comment