The Equation That Destroyed the World
The source of the problem is the rather innocuous equation:
(Portfolio Return)(t) = alpha + beta * (Market Return(t)),
with alpha and beta being constants.
How this equation is used in practice is that we look at the monthly (or quarterly) returns of a portfolio and its benchmark, which is we will assume is an equity index for simplicity. We then run a fit over the time series of these returns, and estimate alpha and beta. The objective is for the portfolio to generate a positive alpha, and high beta is bad.
Following the logic of this equation led to the recent equity market carnage. In addition to some people losing savings that they probably could not afford to lose, financial firms will cease to exist, and their employees will need to find gainful employment elsewhere.
Why Did We Care About This Equation?
There is a simple story behind the equation. Everyone in finance academia agrees that short-term interest rate products underperform other asset classes over time, and it is hard to disagree with that sentiment. Therefore, the easy way to outperform the benchmark is to apply leverage to an equity portfolio. Even if you just hold stocks at the same weight as the index, you will outperform if the equity index outperforms your borrowing cost. Most people agree that this is not cricket, you are supposed to be doing something non-trivial to be be considered a good investment manager.The equation above solves that problem; just applying leverage to the index equity portfolio will create a return pattern that results in no alpha, just a beta greater than 1. Hence, positive alpha good, beta greater than 1 bad.
Why the Blowup?
Rather than doing something sensible -- barring leverage -- the financial industry embraced the equation. The question then arises: how do I generate alpha?Being like an old timer and allocating capital to successful firms would not actually work. Successful firms tend to grow quickly, and are more volatile than the index. Firms that are being run into the ground by incompetent CEO's tend to stable enterprises (until they go bankrupt). The usual justification for the existence of the finance industry is that it allocates capital. However, allocating capital properly is actually a ticket to losing your job.
Instead, we need to reverse engineer the equation. Imagine that the time period is quarterly. What you want to do each quarter is to get exactly the index return, plus a positive constant.
How can we achieve that? It's actually quite easy (assuming it works!).
- You put all of your portfolio capital into an equity index portfolio.
- You sell a 3-month option that expires worthless.
The option premium (as a percentage of the portfolio) is your alpha.
(The fixed income equivalent to this is to lend 3-month money to an entity that does not default on a leveraged basis, or write CDS protection.)
Once everyone does this, the option premium price (implied volatility) collapses. (Or, the credit spread collapses.) So you sell more! It's going to expire worthless, right?
Needless to say, this works until it doesn't. Either that 3-month money starts defaulting on you (2007-2008), or the realised/implied volatility moves against you (2008, 2018).
This "picking up nickels in front of a steamroller" behaviour of the financial industry is baked into the whole alpha/beta cake, so it will keep happening until the adults in the room finally wise up.
This "picking up nickels in front of a steamroller" behaviour of the financial industry is baked into the whole alpha/beta cake, so it will keep happening until the adults in the room finally wise up.
Concluding Remarks
Until the financial industry stops listening to finance academics, capital will continue to get destroyed.
(c) Brian Romanchuk 2018
Brian,
ReplyDeleteIt's much simpler than this. People just go loopy at the top of markets.
Henry Rech
You say " the easy way to outperform the benchmark is to apply leverage to an equity portfolio. Even if you just hold stocks at the same weight as the index, you will outperform if the equity index outperforms your borrowing cost.". I thought that volatility drag would prevent that. If you were to maintain say a 1:1 leverage, then you would have to sell when the market had gone down and buy when it had gone up. That buy-high, sell-low strategy would cause losses that would counteract the difference between the returns of the stock market versus your borrowing cost. Ole Peters proposed that such an effect is what sets the price of the stockmarket over the long term. Basically it is priced at a price that you can't make a long term profit by borrowing money to passively buy stocks. https://www.santafe.edu/research/results/working-papers/stochastic-market-efficiency
ReplyDeleteI am not saying it is a good idea, I am just explaining why people set up the system to limit the use of leverage as a "no brainer" trade.
DeleteThere's no reason that we have to keep a perfectly constant leverage ratio. Just apply the leverage once, and let the position ride. If you rebalance towards your target leverage once every six months, transaction costs will not be a problem. Of course, you will get creamed if your equity goes to zero.
For example, if you have 50% leverage, you are wiped out in the event of a 50% fall in stocks. Realistically speaking, that would be too much leverage (even though 50% leverage would be laughably prudent in rates relative value trading).
But 10% leverage? If the index return is 7%, you're at about 7.7%, and that is way better than most equity investors would do - without really using your brain. That's the sort of thing that "beta" is designed to catch.
This comment has been removed by the author.
DeleteI suppose standard leveraged ETF funds rebalance every day and so suffer a lot of buy-high-sell-low volatility drag whilst investment trusts with gearing typically let the position ride for longer as you describe and so have the leverage ratio wander. In principle a six month rebalancing cycle could though give the worst of all possible worlds if the market was gyrating with that sort of rythm. The rebalance could buy stocks and sell stocks just before the market turned. I wonder though whether it is easier to profit from leverage when a large proportion of returns come from dividends rather than from share price.
DeleteI also think the volatility drag that comes from maintaining a constant leverage ratio is a bit distinct from what is normally meant by transaction costs. Imagine a situation where there were no brokerage fees and perfect liquidity such that if the stock price were steady, it would be possible to trade back and forth any number of times without it costing anything. Even without transaction costs a fund with a constant leverage ratio would still suffer volatility drag in a market where the price went up and down and so the fund had to buy high and sell low so as to maintain a constant leverage ratio.
DeleteI should have written “transaction costs and volatility drag”; cannot edit the comment now.
DeleteIf you have 10-20% leverage, you would need to have a fairly contrived situation for the volatility drag to be significant, particularly if you use some common sense when rebalancing.
In any event, I am not suggesting the investment strategy, just why it’s being barred.
I don't see why your suggestion is that the "sensible" solution is to bar leverage. Leverage use is limited by the interest cost (so the alpha has to include a negative offset) and the risk of the need to deposit collateral at an time when it might be infeasible.
ReplyDeleteFinancial stability reasons. Equity investors are already at the bottom of the capital structure; levering up makes instability worse.
Delete