Thursday, February 12, 2009

Getting Whipsawed!

Whipsaw is a wonderful description of a trade. Specifically you buy some stock, it goes down 20% and you sold out, intending to preserve whatever capital you have left. And the next thing you know, the stock goes up 100%! Shiok right!

Actually the fear of getting whipsawed is so great that it causes a lot of investors to make silly mistakes. But if you think about it, even if you really get whipsawed, it's not a big deal except for the psychological factor. Say you cut loss at 10% and the stock subsequently rallied, so you would have lost just 10%. But if you did not cut loss and stock continues to decline, you will eventually lose maybe 50-60%.

Let's for argument sake, make the example a bit more mathematical. Say you bought a stock at $10 and it plunges to $8. There is a 50% chance that it may rebound 50% to $12 and 50% chance that it plunges another 50% to $4.

So you have two choices now:

Choice 1: If you cut loss, you lose $2

Choice 2: You wait out the storm,
If you get lucky, you make $2, as the stock rise back to $12
If you are damn suay and the stock continue to plunge to $4, you lose $6

Let's assume it's 50:50 between the $2 and -$6, your expected return of not cutting loss is $-2 (0.5*2+0.5*-6), which doesn't make you better off than if you had cut loss. Actually it's probably 70-80% chance that it will go down. Logically thinking stock at $10 which had gone down to $8 should continue to decline bcos something had gone wrong in the first place. So unless a new positive catalyst appears, the stock will not rally.

However the fear of getting whipsawed is so great that it blurs the rational mind. If the stock did rebound and go back to $12, most pple would rather kill themselves than to admit that they only lost $2. This fear of getting whipsawed makes us hold on to our losses longer than we should.

Tuesday, February 03, 2009

Gambler's Ruin Takeaways

Well actually, Gambler's Ruin has more to do with speculating than investing. Nevertheless I think we can learn a few things from Gambler's Ruin.

Btw Kelly's Formula is

% of money = odds of winning - (odds of losing / payout)

Eg. You think Stock A that have a 60% chance of going up 80%
Then % of money = 0.6 - (0.4 / 0.8) = 0.1

ie you should be putting at most 10% of your money in this stock
But take note that ALL the inputs are arbitrary, the odds, the payout.
The formula output is only as good as the inputs.

Ok here are the takeaways:

1. If you simply buy a fixed dollar amt, like $1,000 for every stock you will go broke as time passes, even if the odds are fair. And you will go broke even faster, esp if the odds are against you. (As far as investing is concerned, in most cases, the odds ARE against you).

2. Even if the odds are in your favour, betting the same absolute amt doesn't make sense, you need to apply Kelly's Formula or its variations to optimize returns. This means that you should always decide how much money to invest based on a % of your total amt of money and not an absolute amt. And this % should be decided by the Kelly's Formula or some modifications (half Kelly etc) of it.

3. Building on the previous point, one of the most popular implementation is actually the much talked about rebalancing method used by institutions and shrewd indvidual investors. Say you have 60% in stocks, 30% bonds and 10% cash. You should rebalance your portfolio whenever the ratios are out of whack. Like maybe stocks go up to 70% during boom time, so bring it down back to 60%. This makes sure you buy low and sell high and at the same time mitigate Gambler's Ruin.

Ben Graham, father of value investing, advocates always maintaining a ratio of 50:50 in stocks and bonds (with possible digression to 25:75 or 75:25). In the same line of thought, when the ratios are out, say stocks go from 50% to 80%, then you should bring it back down by selling. This ensures that you buy when it is low and sell when it is high.