Call us predictable, but there are a few financial terms that always capture our attention, and “correlation” is high on that list. This morning Josh Brown over at Reformed Broker brought to our attention this Morningstar article, lamenting the rising correlation between the S&P 500 and various asset classes. According to modern portfolio theory, this is bad news for anyone seeking a diversified portfolio, because highly correlated asset classes will tend to move in unison, pretty much erasing the benefit of diversification in the first place.
While the Morningstar piece does a great job of pointing out the rising correlations among traditional diversifiers, there is one big problem with this analysis – it leaves out managed futures. For comparison, managed futures (as represented by the Newedge CTA index) sports a 1-year correlation to the S&P 500 of -0.28 for March 15, 2002, which decreased (grew closer to zero) to -0.19 for March 15, 2012, giving the asset class a far lower correlation than most of the others on the Morningstar list. Why has managed futures remained mostly non-correlated within a tide of rising correlation?
In our view, it breaks down like this: while real estate and commodities may have been statistically non-correlated in the past, this did not stem entirely from a fundamental difference in price drivers for the different asset classes. In actuality, stocks, real estate, and commodities often rely on the same fundamental factors for gains in price – mainly economic expansion, low interest rates, and the availability of capital (liquid markets). So while some different asset classes like real estate and commodities have been non-correlated in the past, we’re not so sure these lower statistical correlations weren’t due in part to data and price mismatch issues (for example, the real estate wasn’t priced in real time like it is now via ETFs, the commodities were priced via cash instead of the tradable futures, etc.)
On the flip side, managed futures have been statistically non-correlated in the past and remain statistically non-correlated today because they are not reliant on the same pricing factors (sometimes they are, and sometimes they rely on the opposite). That is, they are both fundamentally and statistically non-correlated. Many people only look at statistical correlation, but what really matters is what is driving returns. If two asset classes are essentially reliant on the same inputs to be successful, guess what: they are going to be correlated at some point (especially in a crisis). How do they do this? It’s not rocket science. They achieve this non correlation by going long and short in multiple markets: stocks, commodities, interest rates, and currencies. That short part is the key, as it can make managed futures negatively correlated to other assets when those assets classes are going down due to economic contraction, rising interest rates, or a freeze-up in liquidity.
You might say that hedge funds are in the same boat, because they can and do go long and short stocks, bonds, and so on. However, there are two key differences:
1. Hedge funds for the most part rely on borrowed money to magnify their returns. This puts them at risk of not being able to do the trades they want or facing margin calls when they are not able to borrow short term funds at the terms they require to meet their return/risk targets, i.e. when liquidity dries up (as in 2008). So just like stocks, bonds, and real estate – they are tied to interest rates and liquidity.
2. Hedge funds are at risk from investor sentiment in their own program. A worldwide selling crisis in other assets can result in liquidations among hedge fund clients (to free up cash), which in turn can cause for forced sales of positions from the hedge fund. This can lead to a vicious cycle in which losses for the fund cause further redemptions, leading to more losses. In short, your hedge fund investment is at risk of other people panicking.
Managed futures don’t suffer from these issues. Futures traders leverage via notional funds rather than borrowed funds, which means they don’t have to worry about liquidity in the broader market. In addition, positions are managed per client, so other clients panicking will have no influence on your positions. It may only matter once a decade, if that – but it’s a nice little nugget to help you sleep better at night, to be sure.
Moral of the story? If you’re looking for diversification from these other asset classes, then yes, you should be worried. If you’re in managed futures… well, you figured it out.