Investing can also distort the true risk of loss. Suppose there is a 90% chance that an investment will produce a 10% return each year for the next five years – those are pretty good odds – but perhaps the more important question is how much can you lose if that 1 in 10 chance happens? If the loss is 20% and it occurs in your first year, the actual annualized 5-year gain is 3.2%, even though four out of five years produced 10% returns.
I used 20% because a 20% loss is the definition of a bear stock market, but that’s the starting point and not the limit of a bear market loss. In the last bear, market many stock indexes dropped over 50%! Even if there had been a 90% chance that an investment would return an amazing 19% a year for the next five years, if the first year suffered a 50% decline you’d still have a loss at the end of the five years.
The belief in 2006 was that there would never be a mortgage bond default crisis because the mortgage bond quantitative model showed the odds of this happening were too remote. What they should have done was shut off their computers and asked “what if?” Our model says that the possibility that 100% of our “AAA” rated bond portfolio going into default is very remote, but what if it did happen multiple times and what if these defaults forced one of the largest investment banks on the planet out of business? Could that scenario result in an international banking crisis, and if so, how would that affect the perception of other good quality bonds and the ability of homeowners to borrow? And the answer is: Lehman Brothers did fail and the world narrowly avoided a global depression.
The problem with investments is that it’s impossible to create finite linear math models that can predict the unpredictable. What that means is to truly understand the risks, sometimes you need to throw away the spreadsheets, the charts, and the purported “odds” and simply ask, “what if?” You may decide to go in a completely different direction.