I've often seen the taunt "even a broken clock is right twice a day" leveled at Fin/Econ pundits, especially those whose opinions don't seem to track the stock market very well. I have a hierarchy of clocks. In my hierarchy, a broken clock is the second best kind of clock to be. And if you are an investor, I think you are better off trying to be a broken clock than any other kind. Here's my rundown of clocks:

The Accurate Clock — Obviously, the best kind of clock is one that always tells the time correctly. In stock market metaphor, this refers to the pundit or trader who can accurately time the market, call actionable peaks and lows and be right far more often than she is wrong. An accurate clock can become wealthy at a rate proportional to its accuracy. A perfectly accurate clock could become a billionaire in a single busy day. But, there are no perfectly accurate clocks. It may be impossible to be even a moderately accurate clock.

The Broken Clock — The prototypical broken clock never moves. Its hands point steadfastly towards a single time of the day. Broken clocks have one great virtue: They get to be right twice a day. Always. Every day. They are the turtle to the accurate clock's hare. Rather than chasing the time around all day, they wait for the time to come to them. Among investors, so-called "permabears" or "permabulls" are usually referred to as broken clocks. I don't think it's good to be a "perma" anything. The world changes. A good broken clock is an investor who makes smart decisions based on the best available analysis she has, and then waits to profit, even if circumstances change, even if she comes to doubt her original reasoning. Behavioral finance types refer to this approach with words like "status quo bias" or "endowment effects", like these are bad things. That's okay. Broken clocks probably make more money than behavioral finance types. Being a broken clock is a perfectly sensible strategy in a world where information is worse than imperfect and there will always be great reasons to bail and take losses on positions that will eventually be winners. Note that not all broken clocks are alike. The hands of many broken clocks move, just not at the right speed. Some broken clocks move more slowly than real time. These clocks will always get to be right sometime, but not necessarily twice a day. Then there are the clocks whose hands move fast, or even backwards. These clocks get to be right more often than twice a day! Broken clock investors don't try to just be broken. They compete on the basis of how long they have to be wrong before they get to be right. But broken clocks know that they really have no idea what time it is, and that they'd better be prepared to wait. They avoid inherently time-limited securities like options, and are cautious with leverage, because they hate to be forced to quit while they are wrong.

The Not Broken, Not Accurate Clock — The very worst kind of investor is the clock that keeps perfect time, but it is just not set properly. This kind of clock is never, ever right. It's surprisingly easy to be a not broken, not accurate clock. If a clock is not broken, odds are 11 out of 12 that it will never even get the hour right. Not broken, not accurate clocks are always chasing reality, constantly moving, and never winning. In stock market terms, not broken, not accurate clocks are persuaded that it's reasonable to take losses when circumstance change (as they always do). They take positions for good and sufficient reasons and liquidate them for good and sufficient reasons. In the meantime, they usually fail to profit. Most people who try to be accurate clocks make it half there. They end up moving at precisely the rhythm of the market, but out of phase, buying high into optimism and selling low into despair. Capital markets love not broken, not accurate clocks in the same way you or I love might love pancakes.

It's pretty clear that I'm a big fan of broken clocks. That's self-serving, because I am one. That means I am usually wrong. I'm wrong right now, portfolio-wise. I'm knee deep in red. Whether that fact recommends my advice as authentic, or damns it as stupid, I'll leave for you, dear reader, to decide.

Update History:
  • 08-Oct-2006, 4:16 p.m. EET: Cleaned up some minor wordiness. Fixed place where I said "broken clock" when I mean "Not Broken, Not Accurate Clock".
Steve Randy Waldman — Saturday October 7, 2006 at 10:55pm permalink
Aaron Krowne (mail) (www):
"Hear, hear"... from another broken clock.

What a wonderful analogy! I'll have to steal it someday.

I'd like to think at this point I've built up a fair amount of intuition/wisdom about the nature of markets. And I know enough to know I can only hope to be a broken clock. In fact because of the nature of markets, I don't think accurate clocks can really exist---the system resets itself when they become even slightly successful! Which is why I always chuckle at backtesting---so easy, so objectively flawed. Probably the closest it comes to being an inherently accurate clock is manipulation/intervention...

Anyway, for me, this means that I should do my homework on fundamentals, come up with strong investing theses, and stick to them, no matter how painful it is. The only thing that should move me is information that changes the fundamentals, new insights about the thesis, or (typically) the event of the thesis playing out to completion.

This makes me a broken clock. The market, I am quite convinced, behaves chaotically, modulo fundamental trends. And since there is no agreement on those in advance, the market just looks chaotic. [In fact, there is usually not agreement on fundamentals until long after the fact!]

Now, if you can only figure out how to work inflation into this analogy, maybe we will glean new insights on just how many times a day we need to be right!!!
10.8.2006 3:27am
Steve Randy Waldman (mail) (www):
Aaron, I enjoyed your chaos paper, and am unsurprised to note that we have similar view of these things. (It's worth noting that the grand-daddy of fractal math has a similar view.)

"Probably the closest it comes to being an inherently accurate clock is manipulation/intervention..." I think that's right. It's mainstream to believe that the efficient market hypothesis is not perfect, but basically right in the sense that "there is no free lunch", that one cannot achieve extraordinary returns without assuming extraordinary risks, and that most apparent violations of EMH (small caps, deep value) involve assymetrical downside risk. At the same time, when very large, well connected investment banks achieve very high returns trading for their own account, it is mainstream to chalk that up to "sophistication", rather than manipulation, inside information, or extraordinary risk taking. These two bits of conventional wisdom can't coexist. By virtue of the participation effect that you describe, extraordinarily accurate clocks shouldn't last very long, at least in dollar weighted terms. If they do, the markets they are playing in are by definition far from efficient.
10.8.2006 11:51am
Note: You don't have to register to comment! Just use the pull-down menu to comment as "Guest".