Tuesday, May 27, 2008

Bob Vila Would Not Invest in Fixed Income

Or would he? One of my friends told me he's only 10% in fixed-income in his retirement savings account; for the last two days I've been trying to write a blog post on why that's not a good idea for someone in their 30s, and each time have failed. The problem is I can tell you more about why any investment theory is built on shaky ground than I can about why any particular one is good.

I have some entertaining performance numbers from various investments over the last two years, but while they're good for a laugh, but not much else. If someone says "my investment strategy is to play the lottery" we'd call that person a moron. If the person then won would we go "no, you were smart all along"?

My gut feeling is that my generation doesn't adequately fear the stock market. Consider the following:
  • Periods of relative success increase risk. When things are quiet, we humans with our short-term memory think we can gamble more, until something bad happens. This is essentially what happened with the housing crisis: complex engineered fixed income investments were quiet and steady so people built more and more risk into them for less reward; when they went awry, the result was carnage.
  • Both bonds and stocks have return both on investment fundamentals (the coupon for a bond, dividend and earnings yields on a stock) and speculation (sell the investment on to someone who pays more). Since 1982 we've seen a huge increase in how much people are willing to pay for a stock with a given return. The results of this are two-fold: the investment yields on stocks keep going down, but along the way people make specualtive yield.
  • This means that the potential for future returns keeps going down (since the investment component is getting smaller) and yet we think that stocks do well because their past growth has been high (due to speculative yield). We have to see this for what it is: a pyramid scheme. The speculative component of stoc k yield cannot grow indefinitely, and the investment yield is low due to high prices. I'm not saying stocks won't make money, I'm just saying that when people say "the stock market returns 10%" they're throwing out numbers that were true from 1980-2000 but probably won't be from 2010-2030.
So here is some data, mostly food for thought.

Sometimes the Rules Change

This chart shows the relationship between stock earnings yields, stoc dividend yields, and bond yields. Note what happens in 1955: until then, bond yields have always been lower than stock dividends. In 1955, this relationship changes and it never goes back.

My point here is that sometimes the rules of the game change forever. Benjamin Graham said that in the short term the market is a voting machine, in the long term a weighing machine. Basically he was saying that in the long term speculation will wash out and your long term return will be based on fundamentals.

But I'd say: sometimes in the long term the rules of the game change and you don't win what you thought would be yours. If you made investments under the assumption in 1955 that the weighing machine would vindicate you, you'd never make your money back.

This has applicability to two cases:
  1. More strongly for "regression to the mean" or "value" investors - that is, investors who try to identify temporarily incorrect prices and invest to profit when they come back. This is often a good stragey, but it is not risk free - every now and then the rules change and you blow up.
  2. More subtly, but also more importantly, this also applies to arguments that "X has never happened in the past". For example, you'll find a ton of posts in the last year suggesting an over-weight in equities for retirement investors. The logic is: they return more and over N years they've never lost money" where N is long enough that the stock market recovered. There are specific problems with this argument I'll blog about some other time, but it's worth noting that in 1955 stock dividends had never been lower than bond yields, until that wasn't true forever.
Sometimes the Rules Don't Change

This is a graph of the Nikkei 225, that is, Japan's major stock market index. My generation is too young to really understand Japan in the 1980s I fear. Note that not only has it not made back its peak pricing, it hasn't even come close.

Whenever there is a bubble, you'll hear "this time it's different" - think dot bombs that lose money on every transaction but make it up in volume with $100+ stock valuations. Sometimes the rules change, but a lot of the time they don't.

The Nikkei provides a number of warnings about stocks:
  • If your underlying asset is subject to bubbles and heavy speculation, you might never get your money back. A diversified, balanced portfolio is a good defense to this.
  • When you invest really does matter...if you hear "99% of the time strategy X returned great results" ask yourself: what happened in the 1%? Would you take a drug that will fix your headache 99% of the time and kill you 1% of the time?
  • Stock markets don't "always go up". That's what they said about housing.
The Life of a Turkey

There is a wonderful graph in The Black Swan called "The life of a turkey" - it is a steadily increasing graph of the amount of food fed to a turkey on a given day, until right before Thanksgiving the graph abruptly plummets to zero.

The fundamental problem with a whole pile of the financial tools we rely on is that they look at past data to answer questions about the future. But, as the life of the turkey points out, before Thanksgiving there was no data in the Turkey's history that could predict what was going to happen.

Taking Stock

And this is where I recommend a certain amount of skepticism when approaching stocks. You will see articles like this one, but as you read, remember the life of the turkey! In no 30-day period in the turkey's life did our poor bird die once - and yet he was not immortal.

There is simply no guarantee that trends from past data will sustain themselves - the "30 year win" theory of stock investing does not make a stock investment safe. Stocks can and do lose money - we have to look at the underlying mechanism of the investment, not just what the numbers have done in the past.

Stocks do not become like T-Bills over 30 years just because their average peformance shows the same volatility over a long enough time frame!

To end with my favorite epistemologist, Donald Rumsfeld, there are known unknowns and unknown unknowns. Past stock performance might give us some insight into the known unknowns, but it does not tell us anything about the big unknown unknown: namely how and when have the rules of the game changed?*

Unfortunately we go to war with the investment analysis tools we've got, not the ones we want. I am not saying "put your money under the mattress", and I do believe that diversification is a damned good idea.

My point is this: when you look at "risk", consider whether scenarios exist where you would lose more money than you are comfortable with, not the average and typical outcomes. There is no guarantee that the average or typical will happen, or even that it really is likely. Today Modern Portfolio Theory has become gospel, but don't let the gospel keep you from thinking about the unthinkable.

* Of course, people now are saying the rules have changed because we will have "decoupling" of the US and Asian economies. This strikes me as even more ludicrous than past statements about dot.bomb valuations, but that's another blog post.

No comments: