$34.95

Books have arrived; now shipping

Uncorrelated

Financial engineers or quants have become famous and highly compensated on Wall Street in recent decades. One of the things they supposedly do is manage risk better.

Superficially, the subprime crisis may seem to have destroyed that reputation. Actually, it’s more complex than such a sweeping statement suggests.

But one of their main techniques is to allocate assets to different categories. Indeed, that has been standard financial planner advice for decades: the old don’t put all your eggs in one basket usually expressed like this example from page 34 of David Swensen’s book Unconventional Success.

Domestic equity 30%
Foreign developed equity 15%
Emerging market equity 5%
Real estate 20%
U.S. Treasury bonds 15%
U.S. Treasury Inflation-Protected Securities 15%

Some also suggest 5% or so in gold and other precious metals.

You need to understand that the whole idea of putting your eggs in more than one basket is that the baskets are “uncorrelated” to use Wall Street quant terminology. That means they are unrelated: what happens in one basket does not affect what happens in other baskets. For example, what happens to bonds does not simultaneously happen to stocks.

It turns out that virtually all investments that you do not take physical possession of are pretty correlated. Indeed, even assets you do take physical possession of are somewhat correlated to the organized markets. For example, gold in your safe deposit box is affected by the price of gold as established by the commodity markets.

Where are the correlations?

What connects one basket to another like the stock and bond baskets?

• interest rates
• common ownership
• new laws or regulations that cover more than one asset class
• copy cat imitation of investment strategies
• gigantic amounts of money flowing from one asset class into another that has recently outperformed
• computer trading where the algorithms relate buying and selling one asset class to activity or prices in other asset classes
• common lenders
• supply chain

The Hunts and silver

Here is a classic example from all the way back in the 1970s. Silver went down in price. And so did cattle. What the heck would connect silver and cattle? Sales of spurs and cowboy belt buckles? No. It was the Hunt Brothers. They tried to corner the market in silver. They failed. When the price of silver began to fall, they had so sell other assets to come up with cash. They owned cattle and sold a lot of them to get that cash. Thus trouble in the silver basket spreading to the cattle basket through the millionaire Hunt Brothers owning both assets.

The Hunt Brothers also give an example of new regulations. Some members of the exchange where silver was traded bet against the Hunts on silver then changed the rules of the exchange to hurt the Hunt’s long position and favor their own short position. Of course that hurt not only the Hunts but also every investor who had a long position in silver.

Long Term Capital Management

In the notorious Long Term Capital Management collapse, LTCM CEO John Merriwether thought he had all the risk managed by use of uncorrelated assets. When he went down the tubes, he expressed bafflement as to how the various asset classes got correlated. one of his friends in the business told him, “John, the correlation was you.” Indeed, after extreme initial success, massive amounts of money flowed into LTCM. LTCM expanded into other investment strategies and asset classes. Because they were extremely highly leveraged, any slight adverse movement in one asset forced them to sell other assets to raise cash. The selling of “good” (at the moment) assets to meet margin calls in highly leveraged “bad” (at the moment) asset classes caused contagion between the different assets such that a slight downturn in one resulted in a downturn—caused by large amounts of selling—in others.

Subprime crisis

A similar thing happened in the late 2000s with the subprime crisis. The subprime investments by Wall Street hedge funds were highly leveraged. When the subprime mortgages began to default, the highly leveraged owners of those assets had to sell other “good” (at the moment) assets to meet margin calls, like stocks. That is why the subprime crash caused the stock market crash and so on.

Common lenders

An amazing number of business depend on lines of credit and even overnight borrowing to survive. Add to that the fact that some lenders are monstrous in size. When one group of a big lender’s borrowers fail to pay back teir loans on time, that can cause the lender to pull back to rebuild its capital reserves. That means less lending by that lender. Lending cutbacks by a lender because of defaults among one category of its borrowers can thereby adversely affect other healthy borrowers of tat same lender.

For example, the Asian debt crisis in 1997 hurt big lenders causing them to cutback on loans to other regions in spite of lack of financial troubles in those regions.

Supply contagion

If a lot of companies in, say, the U.S., depend on cheap labor in Asian countries like China, and the currencies of those cheap labor countries rise against the the dollar, the U.S. companies that rely on those cheap imports will be hurt.

Are there ANY uncorrelated assets?

About the only uncorrelated asset classes I can find are hard assets you own for your personal use, your home, and assets you own for investment or business, like stocks and bonds.

What happens in the securities markets, like an increase in interest rates, will affect your home value. But if you own it for a place to live, not to sell, the value of the home is irrelevant to your purpose for owning it. You simply do not care about the value as long as you intend to live in it.

The same is true of all the other hard assets you own for your use including your vehicles and the supplies of computer paper, shampoo, food, clothes, and so on you have in your home.

Normal times versus panic

There is also the issue of are we in normal times or in a panic or period of irrational exuberance.

In normal times, different asset baskets seem to be uncorrelated. In a panic, on the other hand, you see headlines like “flight to safety.” That means everyone sells everything and puts the sale proceeds into the U.S. dollar like U.S. treasury bonds or gold. In that case, everything is correlated and two things, the “safe” asset and all the others are negatively correlated. That means they are like on a see saw. When one goes up, the other goes down.

Similarly, when there is irrational exuberance, as in the late 1990s dot-com boom, large numbers of people dump “safe” assets and buy the hot asset of the moment—Internet stocks in the late 1990s. Again, there are just two baskets: Internet stocks and everything else. And again, those two baskets are see-saw or negatively correlated, not uncorrelated.

There is a web site that purports to tell you the correlations of various asset classes. Seems like the only way they could calculate that would be to look at past prices of the assets in question. But as every prospectus says,

Past performance is not necessarily indicative of future performance.

You do not need complex algorithms to know that when the market crashes, investors flee from the disfavored assets massively and the correlations seems to be “safe” asset versus all the others. Similar in the opposite direction when some asset is “hot.” I do not see that in the table at www.assetcorrelation.com.