Mon 24 Jul 2006
I named this post because of the Peter Principle, which essentially states what the link says it states.? I was reminded of the mention in Running Money of the fund managers who, when they reached a goal for rate of return,?took windfall profits and sat on them so as not to miss their targets for the month.? So my question is, should an investor call it quits during a given timeframe if they hit a good run and, if so, when and for how long?? It’s much like the problem facing?the smart gambler in a casino who knows that the long run spells disaster and so?cashes out after a miracle run of dice.? But the trader is in a wierd spot.? They usually are in the game because they think, in their infinite wisdom,?it is very much not like a casino in that they, not the house,?have?the edge.? So if you all of a sudden have a week where you make enough money in the markets to make for a decent year that even Buffet of Soros would be proud of, should you stop there?? When do you start back up?? I mean aren’t these divisions of time called “quarters” and “years” really just arbitrary as far as the market is concerned (don’t tell your accountant I said that)?? Whose to say you won’t double that return in another week?? Whose to say you won’t lose it all simply based on the reversion to the mean or to your true level of (in)competence?? Even if you do have an edge, its true value may lie somewhere well below the returns you achieved recently, which may have had a healthy does of luck in them.? Normally, you don’t have a benchmark to really tell you where you should be.? In the case of something like the Uberman’s Portfolio, you do know one thing: the amount of the return due to interest vs. directional moves.? So there is a great example.? What would you do if the expected yield from interest for year is estimated at 20-30% and all of a sudden you find your $10,000 account sitting at $12,500 after some lucky market timing (e.g. you rebalance and it just happens to be a major market bottom in your heaviest position)?? You know your model is designed to neutralize market moves but it isn’t perfect so it’s always possible you might catch the right wave (or the wrong one).? This also means that you might revert back to pure interest income any day.? So do you stop right there since you have reached the theoretical annual goal early?? Or do you ride it out and take what you are given in the name of consistency?? You might be on your way to a 100% year, luck?or not,?so why stop?? What about the flipside of all of this, where you experience larger than expected losses?? Is it supposed to make me feel good that most of the sentences in this post are questions?
July 24th, 2006 at 9:55 pm
My first reaction is that it depends on what your objectives are. For example, if you’re trading to earn a living and you just hit your salary for the year — maybe it’s a good time to go on vacation, do some additional research, etc. or scale back your positions to control the risk of giving those profits back.
If your objective is to have a high return within certain risk parameters, then you may have misjudged your risk. Specifically, if “luck” can push you into more profits than anticipated, luck can just as easily push you into losses. Either you misjudged your risk or your misjudged the probability of a rare event. I remember reading one of the Market Wizard interviews (years ago, so it’s a little fuzzy) where a guy had a 35% return one year when his models predicted a max of 25%. He found he had inadvertantly exposed himself to more risk than originally planned. He tuned down his risk and continued trading. (Think about this one in the context of the UP strategy.)
If I were a discretionary trader, I would probably look to see if the excess profits were due to a trade gone right or just luck. Self-delusions aside, I would probably scale back positions (or take some money out) if I deemed the success to be luck.
In general, if you know the parameters of what you’re trading, a large deviation from those parameters is probably worth a step back. If I keep getting blown out of my stop losses, I need to re-evaluate whether or not my strategy is viable and tradable. Same for the upside surprises.
In my current sitation, if I hit my profit targets early, I would probably take a couple of weeks off from trading (at least) to evaluate what happened. Oh, and celebrate.
July 24th, 2006 at 10:10 pm
On closer inspection, I’m not sure if your “peter principle” link the the original post is on purpose or not… it links to http://peter_principle/. If that was on purpose, it might be irony too subtle for me to notice…
July 24th, 2006 at 11:46 pm
Much agreed that one should always monitor their expectations vs. reality. But there is a problem with that guy’s story (granted I don’t know the details). There is no such thing as a maximum. There will always be the outside possibility of anything occuring. Just ask LTCM. If your “max” means 3 SDs that doesn’t mean 4 SDs is impossible, only unlikely. Not sure if I’d label 35% in a year as outside the realm of possibility on a 25% system. What matters is the frequency of those events and whether they tell the trader that a different distribution must be at work. But this you know.
What’s interesting though is if the expected return falls within your risk parameters. What if you expect 20% but you don’t see 25% drawdowns as impossible (rare but not impossible). In that case, the occasional year of hitting your numbers early wouldn’t be out of the question. I guess that’s where your vacation rule comes in and perhaps the risk shouldn’t be allowed to have a high probability of exceeding the return.
One thing to ponder in all this is that if you limit your excess returns for the sake of hitting a target, you might be giving up the cushion that averages out the periods of loss and vice versa. Perhaps you just need to leave the system to work itself out and hold the course while monitoring risk levels carefully. I’m speaking purely of probabilistic models though, not necessarily ones that involve high levels of discretion. There the “working itself out” involves more introspection and planning about one’s next move.
July 24th, 2006 at 11:48 pm
Link fixed. That’s the 2nd time that has happened to me. It’s something about the way I’m dealing with links in the editor that causes them to change the underlying HTML to these generic URLs.