Finding symmetries in an unsymmetrical world ..

Survivorship bias : Asking the uncomfortable questions

Posted on: February 10, 2010

Survivorship bias is the logical error of concentrating on the people or things that “survived” some process and ignoring those that didn’t.In other words , focusing on the victories to mask the wrong that has been done .

That is to say masking the results of poor performing assets/instruments to cosmetically improve the appearance of the lot .That being said,it happens all the while withe the Mutual funds and other Financial Instruments , or with the Top performers of every college being idolized and revered with the Marketing Clinches.

With top performers , any college masks out the droppers, mediocre students who did not get a job  (worser still got a non- value add job – say selling burgers – or documenting- figures don’t account this “averaged” cases )  – Which is to say that Averages is always misleading .What we need to look at is Random numbers , probability of the distribution of the top performers in each of the specialization ,and consequently mapping this with your abilities -and finding out your chances of survival !

Also , finding out the standard deviation from the mean gives a fair estimate of the deviation to be expected , or that which is both normal and acceptable .

Now let us see what that means in game theory,and how could we detect the same ?  Let see this with an example of Money Markets (simply because the payoffs and the utility function are better understood in this context -and there is a measurable set of parameters to evaluate this “bias “ on .)

The rule of thumb in Game Theory ascertains to know the strategy of the other player-and hence predict what would be in his self interest to make the move .

That in an Asset Management organization could mean – to better reflect and sell other “far better performing ” funds  –

a. To close the funds that have higher risk and comparatively low returns .

b. To Mask the data by first implementing (a) ,and then taking an average of the subset of the better-off funds -and declare the weighted Performance average to be far better than it is .

Now let us see the risks – i.e what this means to an individual investor

There are two cases here –

a. That this investor invested in the “bad” funds .

b. That this investor invested in the subset of funds minus the bad funds (i.e the better off funds) – so no impact

* potential impact discussed in point (c)

c. That this is a new investor seeking to distribute his portfolio across the different funds of the AMC .

In case b and c , the misleading data hurts because – it gives an overall impression and hence the incentive for further/future investment ,which harms the investor by increasing the risk,and lowering the returns .

The safeguard – To find what the average actually means – is it a weighted average of all the funds of the AMC , how long have the funds being running since inception, the debt/ equity exposure and hence the risk vs the returns , the brand image of the AMC , the returns quoted against per year performance or averaged over three/five /other periods – mostly the firms “use” 1 year retuns- simply because they look lucrative and better instruments .

Also , these are the things to look at –

What were the market conditions that the performance intends to capture – bull, bearish , averaged markets .

Is the firm comparing apples with apples – Is the comparison between different funds of different AMC ‘s taking the market conditions into account . Is it evaluating it against the same averaged overall period ?

Are the figures exaggerated -what else is potentially beautifying it ? Is the mathematics correct ?



You don’t need newer landscapes to make new discoveries – just a pair of new eyes  – Raffael  Lomas


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Hey there, would be nice to have you around !

That’s me

My live Rantings @Twitter

Blog Stats

  • 13,031 hits
February 2010
« Jan   Apr »

Top Clicks

  • None
%d bloggers like this: