Wednesday, June 21, 2006

Playing It Safe is the Riskiest Strategy

In my post on "what you know can hurt you", I made the point that markets reward the assumption of risk, not the pursuit of the familiar.

Here's another case in point.

Since January, 1997, we've gained 50+ points in SPY, or the equivalent of over 500 S&P points.

So you'd think it was a reasonably good market for daytraders over that time.

Wrong.

Daytraders like to close positions out by the end of the day to avoid overnight risk. But markets don't reward the avoidance of risk.

From the close to the next day's open, from January, 1997 to the present (N = 2402 trading days), SPY gained a total of over 139 points. During that same period, from the day's open to the day's close (the daytrading market), SPY *lost* a total of 88 points.

So, in other words, daytraders who bought the open and sold the close every day during a period of rising market prices lost about 880 S&P points.

But maybe you're thinking this underperformance was just a function of the 2000-2003 bear market.

From 2003 to the present (the most recent bull market), SPY from the close to the next day's open gained a total of almost 36 points. And the daytrading market (open to close)? During the entire recent bull market SPY gained a total of four points. In 2006 thus far, SPY has gained about 6.6 points from the close to the next day's open. During the daytrading market, we've lost about 6.1 points.

It's awfully hard to find *any* recent time period in which the "safe" strategy of closing positions at the end of the day provided traders with opportunity.

Markets are a lot like relationships: Playing it really safe is often the riskiest strategy.

11 comments:

Dave said...

Are you also assuming those traders didn't short the index at any time?

John Wheatcroft said...

Holy smoke - I started the overnight buy - sell cycle several months ago - buy at 3:45 and sell at 9:45 because all of a sudden it was the only way I could make a buck. Actually the time frame is sometimes a bit longer - I learned to wait until 10:15 and then look at the last three 15-minute bars for the DIA if two of three were green I'd hold my trade and continue holding until 2 of the last 3 turned red. Then I'd sell. Made as good a timing indicator as any I could find.

Glad your research has validated my method. Thankyou.

Brett Steenbarger, Ph.D. said...

Good point, Dave. There is research to suggest that daytraders as a whole have a bias to the long side, but you're right: if the traders were shorting the index intraday, they might even have an edge! Thanks for the note--

Brett

Brett Steenbarger, Ph.D. said...

Hi John,

That's an interesting strategy. Seems to me that it has been able to capitalize on the relative strength from the Asian and European averages, which then become reflected in the U.S. open.

Brett

John Wheatcroft said...

A lot of day traders these days are playing both sides of the market. The most successful look for set-ups starting about 30 - 45 minutes after the open and take the play as it lies up or down. Set-ups are found most days in gap-ups/downs. Generally speaking a good set-up is always available. Yesterday for example I caught LNN on the 15-minute bars. Made a good entry and clean exit for a few fun bucks.

I play every once in awhile but it is too intense for me to do more than a couple of times a month. My ADD kicks in and I need to do other things than watch a trade develop. I just do it for fun and to try to keep the mind sharp.

Brett Steenbarger, Ph.D. said...

Yes, good point re: playing both sides of the market. Real-time screening tools (the Trade Ideas program comes readily to mind) can be quite helpful in monitoring setups as they occur.

Brett

longnshortofit said...

Few average investors and most traders are adept enough consistently to keep ahead of the game.

Taking the kids to baseball and other endeavors make it a tad hard.

For the record I think we have a total reversal on the indices today and go higher...Dow and Nasdaq look healthy at about 11042 and 2125...yesterdays close will be tough but I think we get it.

Brett Steenbarger, Ph.D. said...

Hi,

I think you have an excellent point. It is difficult to immerse ourselves in markets and market patterns if we also have to hold a full time job, attend to family needs, and so on. There are very few performance fields that allow one to develop expertise as a part-timer.

Brett

Dave said...

Brett: I'd say that most traders/investors prefer the long side, and it's not just day traders. I think if anything day traders are probably more comfortable exploiting the short side as day trading and short selling both involve a higher risk and higher reward.

Also, in regards to the comments to be immensed in this market. I agree there are alot of things out there that we have to do on a daily basis and to follow every tick or every stock market mover of the day is hard, and although this is a different time from when Darvas invested. I believe if he can do it, so can you. It does come down to waning to take the time to do so, a top all the qualities needed to become a sucessful trader/investor.

Brett Steenbarger, Ph.D. said...

Hi Dave,

One tool that I think helps traders improve their performance is simulation capability. The ability to save market data, replay market days, and practice trading those markets allows traders to greatly accelerate their exposure to market patterns and hasten their learning curves. The NeoTicker (www.tickquest.com) and Ninja Trader (www.ninjatrader.com) platforms stand out in this regard.

Brett

Eddie Offermann said...

Going to start following your blog thanks to this old post.

I'd developed a set of end-of-day metrics that was generating excellent and dare I say "reliable" gains in early morning trading. After running the model for a week and pouring over the charts for the stocks I'd selected from it, I noted a pattern: if I could get in before the close on these stocks, my gain would have been even more substantial.

Since the method involves data that's normally not published until end-of-day, that required some programming this weekend: but now I'm able to generate the same metrics dynamically from live data.

As a trading pattern, I've rarely heard of it: but it was interesting to read someone with long-term analysis to back it up as I've only been following the trend directly for a few weeks.