Blog

Curve-fitting (noun): the act of adding rules to the Key Idea of a trading system for the specific purpose of eliminating bad trades. Systems that have been ‘curve-fit’ never hold up in live trading.

I began to pose this question to myself this week as I surveyed the wreckage of yet another of my “Holy Grail” stock trading strategies:  a Linear Regression-based algorithm discovered through my trading software, thoroughly back-tested, all bias removed, paper-traded for two weeks.  Yes, if I had a dollar for every one of these systems I’ve invented over the past 4 years, I’d be a millionaire by now . . .

Reading a web article by noted trading strategist Henry Carstens titled “Introduction to Testing Trading Ideas” got me thinking about the whole notion of back-testing ideas in the first place.  Carstens goes through the entire ABC’s of software testing:  (1) Choose a testing platform, (2) Identify the Key Idea, (3) Run the test and interpret the results, (4) Pass/Fail the “consistency test”, (5) Test stops and additional entry rules, (6) Avoid pitfalls (slippage, commissions, fills, curve-fitting).  As an article of its kind, I suppose it’s informative and helpful. 

Carstens believes that his process, if followed carefully, can yield excellent stock trading strategies.  The only place where Carstens thinks you have to be extra careful is Step #6: don’t fall prey to adding on rules to improve the results (i.e., curve-fitting).  But here’s the problem I have with the entire back-testing procedure.  It is all the way back at Step #3 where the curve-fitting initially occurs!  Think about it.  A system can’t possibly get a passing grade and go forward to implementation without bias occurring at this very point.  Have you ever heard of someone getting bad results from back-testing and thinking “Yeah, that’s the strategy for me, let’s go live with this loser!”  Of course not.

We humans are pre-programmed to be compelled by what works, not what loses.  A backtest can only expose some successful pattern THAT OCCURRED IN THE PAST.  Our brains will automatically avoid any pattern that is not successful.  Oh, perhaps some of us are more honest with ourselves and will deliberately expand the universe of stocks to play in the future well beyond the select group that initially fit the back-tested pattern.  That’s what Carstens means by passing the consistency test:  “A system that has been consistent over time is far more likely to continue being successful than one that has a history of ups and downs.”  Expanding our universe of stock plays will lower the system’s historical results but ostensibly iron out the hills and valleys and make the system more believable as one that can stand the test of time. 

If only this were the case!  A system deliberately made more consistent is still just another pattern that worked in the past.  Back-testing a system can predict nothing about forward-testing (i.e., playing with real money) the same system, even if the pitfalls are avoided perfectly and the system is played out in real life as mechanically as the software program was able to do. 

Trader software is overall a boon for the industry, and I still use my Amibroker on a daily basis.  But I see more and more that the successful traders are ones who combine mechanical and discretionary approaches.  They do not just let the software blindly lead them along, and they are no longer naïve enough to believe that past results are any indication of future returns.  Technical traders should not be seduced into thinking that brilliant software technology can do the trading for them.  Instead, they need to begin thinking “out of the box” – look through the successful patterns and play with the underlying algorithms in ways hitherto unknown.

I decided toward the end of the week that posting the holding rationale of every Tom, Dick and Harry stock that I happen to purchase isn’t going to be a very fruitful endeavor. While I might eventually get to the top of the Commentary Ranking list, I’m fairly certain that the Covestor community wouldn’t give me high marks for being insightful. That’s because I’m purely a technical trader, couldn’t care less about general market direction, and use two very obscure algorithms to pick my stocks. As I told someone else in the Covestor community this past week, once an algorithm generates a signal, I don’t try to understand the why or wherefore but simply take the trade. It’s worked very well for me in the past and I don’t see any reason to change things in the future.

Certainly there are many technical traders who could write an entertaining rationale about their stock picks. These traders would be in the category of "chart readers" – those looking for patterns such as head and shoulders, price gaps, triangles and wedges, support/resistance, supply/demand, etc. Good chart readers have a lot to say about why they take a trade, and the ones who are at the same time good writers (such as our own Tim Sykes) can really hold your attention. But someone like me – who uses mechanical algorithms with software code that I do not wish to reveal to the general public – isn’t going to find much to write about.

In fact, trying to write an entertaining rationale is what got me in trouble this past week. I began to dig more deeply into companies after I’d bought their stock: earnings news, press releases, industry profiles, analyst estimates and the like. Then I blogged about them. Knowing these kinds of things, or rather dwelling on these things, is the absolute death to my kind of trading. The very success I’ve had is because I know NOTHING about the company – a trade is just a black box, not transparent and not alive. But last week I began to research too much and that led me to begin REACHING too much for a trade instead of holding my ground and requiring a trade to meet my standard or else; and worst of all, I began to listen into a few chat rooms to pick up clues for "good" trades. Because of that, I made a few trades that had nothing to do with my algorithms and I had my first drawdown at Covestor because of it.

From now on, no more blogging about individual picks. I’m going to write a weekly summary of my results from now on, maybe say something about one or two of the trades, and that’s all. If I have an inspiration to talk about something more general, I may occasionally blog mid-week.

That being said, I don’t think 9% gain in my portfolio since March 28 is too shabby. It’s a good start and I’m going to be looking this next week to "get back to the basics," stop reaching for trades and take what the market gives me. Thanks for reading!

Amkor Technology signaled a "buy" this afternoon (Linear Regression algorithm scan).  Purchase price is $9.50.

Pacific Ethanol seems to be doing this daily dance of late bouncing between $4.25 (a few cents from all-time low) and the 15 DMA at around $4.50.  I'm grabbing it at $4.29 - targeting a 4% gain.

I've been impressed that Cal-Maine, the egg producer, didn't psychologically relinquish the $30 mark today.  Food inflation is certainly helping Cal-Maine, which recently reported record revenue.  Since egg prices have gone up 45% in just 8 months (now selling at $2.18 a dozen), Cal-Maine expects to make $6 per share in profit during 2008. This translates to a $2 dividend per share, yielding 5.3% (for those of you interested in dividends).  The stock is down around 22% from its recent high on March 28th.  I’m a buyer here at $29.77.

Starent Networks just threw an oversold Doji candle at me, and I think I’ll respond with an immediate “buy”.  STAR is one of a handful of Networking Communications stocks up on the day in a sector that’s really taking a hit – that’s impressive in its own right.  Also, STAR is reporting earnings at the end of the month and the company usually meets or exceeds expectations (not that I'll still be around when they report!).

NxStage Medical certainly got the stuffing beaten out of it after posting weak 4th Quarter results two months ago.  After hitting the bottom at $3.61, the stock seems to be making a solid comeback.  I like the way the stock has held $6.00 today amid all the selling (it traded as high as $6.74), and it's also part of that medical device group whose performance prompted me to buy OFIX earlier today.  I bought some at $6.04.

Medical device makers (NXTM, ZOLL) are hot this morning, representing some nice sector rotation into an industry that historically does well in recessions.  Orthofix is another beaten down device maker that came up on one of my oversold scans two days ago - I've been watching it, biding my time, and caught it this morning on a dip toward $32.

Update 4/10, noon:  Sold all shares for 1.5% profit.  Didn't like the action on a day like today, and there was some slightly bad news on the stock this morning.

Pep Boys is another favorite of mine, as I've made some nice money on this stock over the last couple years.  The stock has fallen around 20% since it reported bad earnings on March 25, but it looks to be bottoming out - perhaps Pep Boys Chairman William Leonard thinks so too, as he just bought 25,000 shares on Thursday.  This "buy" might be on the risky side (the Oscp/Candlestick signal for this play wasn't particularly strong), but there have been precious few oversold stocks appearing on my screen of late.

Update 4/8, after-hours:  I was in business meetings all morning (yeah, the day job that pays all the bills), so didn't see the Hold coverage initiated on Pep Boys by Stifel Nicolaus.  Had I seen it, I probably would have sold my position right at the open and come away with a 1% gain.  But even though the stock was down 3% this afternoon, Pep Boys seems to have survived the selling, and Manny & Moe may actually have some life in them to play another day!

4/9: Hit my stop (April 3 low).

Acorda Therapeutics, which markets drugs for multiple schlerosis and other nerve-related ailments, came up on my oversold Price Oscillator screen today.  Acorda seems to have pretty strong support at $18.40, which is the price I got it for, and it's currently trading at a market capitalization of less than $600 million, which Motley Fool thinks makes it appealingly cheap.  I've made good money on this stock in the past - when it bounces, it really can run up fast.

Update 4/8, 11:40 - Might have sold prematurely, but a 3.5% profit isn't anything to sneeze at.  I'm out all 225 shares.