Hindsight Bias

By the way, Happy New Year, everyone! Today I will discuss a cognitive fallacy capable of ensnaring us all: hindsight bias.

Here is a forum post from July 2015:

      > Why would anyone have their portfolio weighted with long puts in
      > a major downtrend preceding 9/11? It makes no sense. Karen said,
      > “stick with the trend.” A proper portfolio preceding 9/11 would
      > have been selling calls on each rally and EXTREMELY LIGHT on the
      > put side, if at all. Just look at the monthly chart; look at the
      > size of the candles each selling month. Anyone selling naked puts
      > in this scenario needs to look at the bigger picture first and
      > understand the context of the current market they are in.

I responded as follows:

      > I mean no personal insult but I think this is a very ignorant post.
      >
      > I claim ignorance because hindsight makes it easy to determine
      > trend. How likely is your approach likely to work in the future?
      >
      > I don’t know what rules you suggest for use on a monthly chart
      > but simply using a monthly chart means the number of occurrences
      > is limited from the outset. If you don’t have a sufficiently
      > large sample size then your approach may fall prey to curve-
      > fitting: works in the past but not in the future.
      >
      > Discretionary guidelines can always be bent to fit pre-existing
      > biases. What you have described (e.g. “monthly charts,” “size of
      > candles”) is very nonspecific.
      >
      > Being specific means objective definition of rules. Only then
      > can you backtest and begin validation. Author Kevin Davey
      > suggests a good system may be found for every 100-200 tested.
      > Do not believe it so easy to come up with a set of technical
      > criteria capable of predicting the next big fall.
      >
      > You may say, “I can be wrong an extra time or two as long as
      > I’m out before the fall.” I think this is a good point but
      > realize that whipsaw losses can be significant and this may
      > render adherence to your system very difficult. When it
      > comes to losses, traders can be a fickle lot.

Hindsight bias [which also reminds me of future leak] is a common logical fallacy that must be recognized in trading and investing discussion/literature. The fallacy underlies artificially inflated performance claims. To stay on the path of consistent profitability, our challenge is to debunk such fiction and to minimize its consumption of our precious time and resources.