Friday, May 14, 2010

Hindsight Bias - Don't Watch Your Behind Too Much!

Things are so obvious. Its obvious that valuations in late 2007 were crazy. It was a bubble for sure and one just had to sell at that point. 
Its also obvious that valuations in early 2009 were equally crazy. One just had to buy at that point. 
These market tops and bottoms are so obvious. One really wonders why most of the market participants didn't act on it...
Well!! If investing was so easy, there would be no other profession! :-) All these facts I mentioned above are obvious today. They were not so obvious at that time. In fact, everything that has happened in the past (not just in investing) seems pretty much obvious when we think of it at present. We sometimes lament at obvious mistakes made in the past, or sometimes justify them, so that we don't have to lament! Think about it, isn't that right? This is the basic funda of the concept of 'hindsight bias'.
Wikipedia defines hindsight bias as 'inclination to see events that have occurred as more predictable than they in fact were before they took place.' Boy, thats one big sentence. In essence hindsight bias is the phenomenon where we look at events in the past, and  convince ourselves that we knew they were going to happen and were prepared for it. While in fact, we just weren't. 


I like parties. Parties are fun and can be useful too. If I get surrounded by people talking about stocks and offering me tips, its probably time to sell, whereas if everyone is avoiding me, its probably time to buy. (Well, to be honest, I go to parties for the booze, not to take buy/sell decisions!). Anyway, the following is a very common conversation, which happened last year. We all must have heard such stuff at parties and social gatherings.


Mr.A: Boss, bazaar dekha? Its gone real bad man..Sensex below 10000!
Mr.B: Of course, what did you expect yaar? You really believed in the 'India shining story' and all? It was so obvious that the crash was going to happen. In fact, I had told my friend Mr.C in December 2007 itself, that dude, sell off. We can't sustain these crazy levels. And see what happened..
Sounds familiar? Such dialogues like 'maine bola tha', 'I had told you that time only', etc are extremely common, specially in equity investing. 
Unfortunately, since I am also part of the conversation, I ask Mr.B: Sirjee, since you knew of the crash so confidently before it happened, did you sell your stocks? Or better still, did you go short on the index? I am sure you must have earned a truck-load of money. 2008 must have been the best year of your life!
Mr.B: Umm, wellll actually I did not sell my stocks. Umm, you see I am a long term investor. The losses at present are just 'paper losses' and i am sure that the prices of my stocks will rise a lot again. So its quite ok since I think long term.
Me: Well, to quote Keynes; 'in the long run, we are all dead!'
Mr.B laughs awkwardly and drifts away. Thank god!
This is a great example of how we think about past events at present. About how we justify our past actions to convince ourselves that we were right.
Another good example is that of technical analysts on TV. Now i do not have any disrespect towards any person. Neither am i saying that technical analysis does not work. It may work for some people. I am not one of those. 
This is what is typically said in the morning, before the market opens:
Nifty closed at 5155 yesterday. We remain cautiously optimistic (my favourite market term) on the market. On the higher side, nifty would face resistance at 5182. If that resistance is overcome, nifty could shoot up to 5200, which would be a major resistance. Beyond 5200, nifty would face resistance at 5215 and at 5032. On the lower side, support is seen at 5140. If this support is broken, a very strong support is seen at 5120, below which nifty has a support at 5100. In case 5100 is broken, nifty could free-fall to 5075. We would see strong support coming in at 5060.’

WHAT EXACTLY SHOULD ONE INFER FROM THIS?


Anyway, in the evening, after the market is over, the following is said:
As expected, nifty made an upmove to reach 5182 levels. Overcoming this resistance, it shot up to 5210, beyond which profit-booking set in. Nifty fell to 5150, which acted as a major support, but in the end-of-the-day’s trade, the bears took the nifty down to the support of 5120, to close at 5125.
Perfect post-event analysis, which is essentially not value-adding. Technical analysts are a great example of hindsight bias. (Again, no disrespect meant, these are just the facts)
In the market, everyone has an opinion and everyone is expert at analysing things beautifully after they happen. And everyone has that goody goody mushy mushy feeling that we knew it all along. Well it ain't so!
Now that we know that, maybe unknowingly, we can get exposed to hindsight bias, lets see what damage it can do to an investor;

  1. We don't learn from our mistakes: One of the biggest aspects of investing is learning from mistakes, so that they are not repeated. But those affected by hindsight bias think that they never make a mistake! So the learning gets stunted. Not a good position to be in..
  2. Gives a false sense of security: Investors affected with hindsight bias think that they have successfully predicted and knew all the happenings in the market. Hence, they remain blissfully confident that they will be able to do this in future too. In reality, this might not happen and can lead to significant losses.
So now that we know about this phenomenon, maybe next time, we'll all introspect before we claim to have predicted something that has already happened before it had happened. 
To avoid being affected by hindsight bias, we must all be honest, introspect, acknowledge our mistakes sportingly and let bygones be bygones.

Cheers and happy investing!

P.S. My apologies for the cheesy title.. :-)

4 comments:

Ninad Kunder said...

Hi Neeraj

It is more difficult to register hindsight bias when things go right as opposed to when things go wrong.

One predicts a x return for a stock and if it generates 2x or 3x return due to some external variable, it becomes very difficult at that point to overcome hindsight bias.

Cheers

Ninad

Neeraj Marathe said...

Hi Ninad,
I think that is very very correct..i should have covered that point too in the post..thnx for the addition..
we are all fooled by randomness many a time!
cheers!
Neeraj

Shankar Nath said...

Hi Neeraj,

People tend to follow the mob due to 'social conformity'. This has been displayed time and again in numerous experiments. I wrote a recent post on this, taking Milgram's experiments in 1961 as an example.

http://scrip-tures.blogspot.com/2010/05/milgram-experiments.html

Regards
Shankar

Neeraj Marathe said...

Hi Shankar,
Nice to see your blog. This is truly a fascinating subject, isnt it?..
Yes, i have read about this experiment in Predictably Irrational, if i am not mistaken..quite a revelation indeed..
Cheers and do keep contributing..
Neeraj