Wednesday, October 28, 2015

Survivorship Bias



Survivorship bias is a type of selection bias.
Survivorship bias, or survival bias, is the logical error of concentrating on the people or things that "survived" some process and inadvertently overlooking those that did not because of their lack of visibility. This can lead to false conclusions in several different ways. The survivors may be actual people, as in a medical study, or could be companies or research subjects or applicants for a job, or anything that must make it past some selection process to be considered further.
Survivorship bias can lead to overly optimistic beliefs because failures are ignored, such as when companies that no longer exist are excluded from analyses of financial performance. It can also lead to the false belief that the successes in a group have some special property, rather than just coincidence. For example, if three of the five students with the best college grades went to the same high school, that can lead one to believe that the high school must offer an excellent education. This could be true, but the question cannot be answered without looking at the grades of all the other students from that high school, not just the ones who "survived" the top-five selection process.
Survivorship Bias in Equity Market
"If you had bought Praj Industries at Rs. 2 in 2003 for just 100,000 rupees, it would be worth Rs. 40 lakh today!"
We hear something like this often. If you bought Infosys in their IPO in 1993, you would make a gazillion rupees, so much that you would be reading this article in your holiday home in the Bahamas. Yet, in the early 90s , IT stocks weren't the darling of the markets - more popular was a stock called Arvind Mills, which still exists and has done fairly well for itself; but if you had plonked your hard earned money into Arvind, you would have seen a stock going from Rs. 400 down to a relatively meager Rs. 43 and now back to just around Rs 280. And that was still lucky; a number of other stocks simply went to zero.
What happened? We counted the winners. If you count only the survivors, no accident has casualties.
Survivorship bias is a classic problem with benefit of hindsight. It teaches us important lessons.
1.That what is visible isn't everything .
2.That it's critical to diversify.
For a person investing equally in a number of stocks from stock market darlings to pariahs, the losses on one set of stocks could have been made up by gains in others; remember that you can't lose more than 100%, but you can make many multiples of your investment when you win ("multi-baggers"). If you research stocks well, chances are you will have a few multi-baggers and losers - sometimes many more losers, but the winners more than make up for the losses. The winners become the survivors and investors become geniuses for discovering them; the real genius though was the diversification.
A simple rule in the trading world is to cut your losses and let your winners run.
That's counter-intuitive; many investors believe the exact opposite - that they would book profits when a stock gains a certain amount, say 10%. But should the stock fall, they will wait till they get out at "break-even", even if that takes an enormous amount of time. The problem here is that you limit your gains, and set yourself up for much higher losses - which over a longer term will lose money, even if this strategy works like a charm in markets that go up. When it doesn't work, everyone else is losing money, so one doesn't feel quite as bad. And here, survivorship bias works again -because people who win on this strategy will brag about it, but someone that lost money will probably have sworn off stocks forever.
Imagine you get a letter saying "The Sensex will go up this week". And it does. Another letter arrives the following week saying, "This week, the Sensex will fall". And remarkably, the index does fall. A few weeks of incredibly accurate information then ensues, and after six weeks you are told that further information will no longer be free; you have to pay Rs. 100,000 to get access to the "proprietary black box system". Well worth the money, you think. You pay, and find the system highly inaccurate in subsequent weeks. What happened?
The modus-operandi was to first gather 10,000 addresses of potential suckers, and to send 5,000 of them a letter saying the Sensex would go up, and the remaining a letter saying it would go down. If the Sensex did go up, they ignored the latter, and divided the "winning" 5,000 people into two to run the mails again. After six weeks, they had 150 people for whom they were incredibly accurate; and who are most likely to pay up. A good 2/3rd of this group might pay, not aware of this scam. The tricksters make a Rs. 1 crore killing from this group through a system that is about as scientific as a coin flip, and all for the cost of some postage. All the scamsters had to ensure was that they chose people who didn't talk to each other!
Survivor bias also impacts our views of what makes people successful. "Work hard, and believe in yourself", they say, "look at Narayana Murthy". But what of the millions that worked hard, but didn't make it?
"Warren Buffett is a great investor, you should buy and hold forever, like he does"
"The best business is the restaurant business. See, every restaurant today makes money."
All of the above is valid but again, not all who live by those precepts have become successful. At some point, luck and fortune play a part, a much greater part perhaps than skill itself.

No comments:

Post a Comment