Perceived Risk vs. Actual Risk

We often see debates about whether or not volatility of returns is a good measure of risk.  This debate is related to what I think is a bigger issue: the difference between perceived risk and actual risk.  Perceived risk is influenced by observations and “dollar bias,” but actual risk comes from the full range of what might happen and its influence on buying power.

Dollar bias and buying power

In some contexts we forget about inflation and view dollars as constant over time.  For example, we tend to focus on nominal returns and think that it’s okay to spend gains as long as we leave the principal intact.  But the principal will erode with inflation if we spend all the nominal gains.

Another context where we see this bias is with mortgages.  We can calculate that with a 30-year $400,000 mortgage at 4%, the first year’s payments will only reduce the principal by about $7000.  But even with only 2% inflation, the buying power of the principal will erode by about $8000, and the fixed payments will become easier to make with rising salaries.  Homeowners are making more progress than they think.  If they can keep up the payments, inflation will eventually take care of both the principal and fixed payments.

Observations and what could have happened

It’s natural to be most worried about the things that we’ve seen happen, but there are many more things that could have happened.  Just because some employees invested everything they had in their employer’s stock and it worked out well doesn't mean that it was a good idea.  If the employer had stumbled, the employees might have lost their jobs and all their savings at the same time.

The way we measure volatility of returns is often by looking at past returns over some period like a decade and calculating their standard deviation.  But this doesn’t capture what might have happened.  Measured volatility might reflect actual risk some of the time, but we’re guaranteed to have quiet periods with low measured volatility even though actual risk remains higher.

We see this at casinos all the time.  Craps tables sometimes appear to be “hot” with everyone making money, but in reality, the odds never change.  It’s not safer to gamble at craps when a table has been hot for a while.

Any investment strategy that tries to optimize leverage using measured volatility of past returns is destined to blow up after a period where measured volatility is much lower than actual risk.  This fact played a role in both the implosion of Long-Term Capital Management (LTCM) in 1998 and in the Global Financial Crisis of 2008.

Nassim Taleb’s parable of the turkey nicely illustrates the big difference between past experience and what could happen in the future.  A turkey might think that life is wonderful with all of its needs being met.  It never anticipates that fateful day when it becomes someone's dinner.

Lessons

Actual risk is what might happen to the buying power of our savings.  It is not just what we’ve observed happen to our nominal savings in the past.  Here are some lessons we can take from these facts:

1. Focus on buying power, not dollars.  The main way we get into trouble with dollar bias in investing is when we think long-term bonds are safe because they preserve principal.  Over long periods, inflation can be devastating, particularly when it rises unpredictably.  Long-term bonds are much riskier than they appear.
2. There are risks out there that we can’t anticipate.  Whatever level of risk you decide is right for you based on the risks you can anticipate, it’s likely that you’d be better off with a little less risk.  This line of reasoning is often used to tell people to shift their portfolios a little away from stocks and more to bonds, but remember that long-term bonds are risky.  Sometimes the best way we can deal with unknown financial risks is to save a little more.
3. Any financial plan that adapts to measured past return volatility is likely flawed.  If you’re into the weeds thinking about the Kelly criterion and Sharpe ratio of your portfolio, you’re probably on the wrong path.