Asset Allocation for Fun and Profit

In my previous post I suggested that commodities speculation was a poor idea and not a good use of retirement funds. But the fact that I think it's a lousy idea doesn't explain why a brokerage like Fidelity would suggest such a strategy. To understand why commodities are being promoted as an "asset class" (and Fidelity is neither the only nor the first brokerage to start pedaling commodities as an asset class) you have to look at what half a century of efficient market theory has done to investing.

Modern Portfolio Theory

To understand anything a brokerage ever does, you have to understand at least the basics of Modern Portfolio Theory. Here's the short version, it's not very complicated.

• Markets are efficient - you can't beat them because they already "know" everything you do. So picking stocks isn't going to get you a better retirement fund. You might as well buy an index of all stocks and keep the fees low.
• If two separate markets have low covariance (that is, they go to hell at different times) that's the only free lunch you'll ever see. By investing in both, you can limit how far down your portfolio tanks. If stocks and bonds go to hell at different times, then at any one time maybe only some of your money is hosed.
• Combine these and you realize something strange: the important question is what types of things you invest in (the "asset class") and in what ratio. The important thing is not the details. In other words, your ratio of stocks to bonds matters a lot, but which stocks you buy doesn't.

MPT dominates the "buy" side of investing (e.g. the institutions that manage people's retirement money) in a huge way. "Lifecycle" funds are just funds that adjust the ratio of stocks vs. bonds as you get older, based on MPT.

Go Off the Cliff With the Herd

There's another boot-print from quantitative finance on the face of investment management. Since the market is "efficient", it's not the place of your investment adviser to try to beat the market. And MPT provides cover for this. If I am in my 30s and my adviser suggests that I should be 70% in stocks, and then the stock market loses half of its value, that's not negligence, incompetence, or a breach of fiduciary duty. MPT says that I have to be heavily in stocks for my risk-return profile. And that investment adviser has a certain amount of cover: pretty much every other investment adviser has suggested the same thing, so "no one could have known".

Unfortunately, MPT strikes me as a difficult approach to long-term investing. MPT "works" based on the historical long term (or sort-of-long-term) relationships between asset classes. But because everyone now uses MPT, the "good ideas" that MPT suggests may not be so good any more since those ideas move markets.

(Put another way: if everyone piles into stocks, the return on stocks isn't going to be very good. But MPT says we should pile into stocks because past returns were good. We saw the bull market of the 80s, we piled in, and in the process we guarantee that we won't have a bull market like that again.)

Look, Fees!

Now here's the problem with MPT from the perspective of a brokerage: it makes it really hard to make any money. If I manage your money, I make my return based on a fee (usually a percent of assets under management for a mutual fund or ETF). But the actual asset classes that I can manage keep moving toward lower fee structures. For example, Van Guard's total market return stock index fund (TMI - this is the ETF I use quite frequently) has an expense ratio of 0.07%. And now (finally) you can get bond index funds with low expense ratios too. BND, another Van Guard ETF has an expense ratio of 0.12%. (This is a huge improvement over the 0.5% you'd pay on bond mutual funds; with only a few percentage of return a 0.5% fee was a huge chunk out of returns.)

Fidelity seems to now be charging me $8 to buy/sell ETFs. So let's review: Fidelity takes $8 to make my purchase every now and then, and Van Guard gets between 0.12% and 0.07& in fees. How does anyone make any money off of this?

To make matters worse, Fidelity can't (and no one else can) credibly come to me and say "you should pay more for this proprietary product" - some magic 'beat the market fund' or the advice to 'stay out of stocks'. MPT says that they can't beat the market and they shouldn't be messing with my asset allocation. So no way to add value there. (Heck, if they did come up with a product that went against MPT and it tanked, they'd probably get sued.)

And now we have enough pieces to understand my cynical view of commodities investing. Currently "buy side" money management companies have only two asset classes to sell (stocks and bonds) and the margins on those products have gotten very, very low. But if they can invent a new "asset class", investors would nearly have a mandate to buy it (due to MPT) and while the market is fresh and new, there's still a hope of raking in fees.

A quick screen of commodities ETFs shows expense ratios from 0.3 to over 1%. And ETFs are usually the cheapest investment vehicle available to retail investors.

In summary: it has become to save for retirement very cheaply. Commodities investing isn't necessarily a good idea, but because it's a new market it's still profitable. Investment advisers have to make the case that commodities are an "asset class" in order to justify selling commodities funds to investors.

In all of this discussion I have not mentioned gold. That is going to have to be the last (and separate) part of this rant. Buying oil, coal, grain, or copper - that's commodities speculation. But buying gold is something very different, and in my opinion, even more loopy.