nisiprius wrote:That's the rub. It's the problem with almost all backtesting. It's only statistically valid if you do it once. If you do it over and over again with many strategies, it means nothing.
If you test strategy after strategy after strategy, sooner or later you will come across one that seems to work even though it's by chance alone. If you test 1,000 strategies, then show someone the best, they'll judge "wow, there's only one chance in a thousand you could get that just by luck." It's worse if you start out with "successful" published strategies, because these are all ones that are known to have worked in the past. And of course it's hopeless if you do what just about everyone does: try something--pick stocks that have a florble-to-gnorgl ratio of 3.62 or better, notice that twenty of them outperformed but the other two were such stinkers that they lost more than the other 20 gained, notice that the two stinkers both had more than one vowel in their ticker symbols, and create a revised rule: stocks with a florble-to-gnorgl ratio of 3.62 AND no more than one vowel in their symbol.
Well put, and I definitely concur with the sentiment. It all goes back to the addage, "statistics don't lie, but statisticians do." I'd say that this specific strategy does fall under the florble-to-gnorgl ratio issue, although if you have a baseline strategy that is at-or-above market performance and then add in your florble metrics, you're still essentially executing the baseline strategy through rose colored glasses. I will contend that a strategy which is constantly tweaked many times over the same set of data actually does
mean something, however. Go back to the Deep Blue experiment of 1996. It had a database of games and a logic based strategy method that it executed to try to beat those games. Many iterations were attempted, but the end result was still the first ever world chess champion beating computer. Perhaps it's not applicable to this situation, but be careful in the generalization of the approach.
staythecourse wrote:Kudos for your friend to refer you here. Many of us have our biases (like anywhere else), but you will not get more informative, level headed advice anywhere else.
To your question... I don't have the answer nor do I care. You will find the more you learn about investing the more it has NOTHING to do with what folks write about in books. One thing I can say is look how easy it was for you to get all that information so you gotta figure how many folks have already tried this method with the same data and none have come out ahead of the others. The author of the book is the same man who wrote extensively about the Dogs of the Dow and that never panned out for any investor I know. The Motley Fool folks found that out the hard way. What makes one think this new method (if it is much different) will do any better??
If you are interested in investing I would focus on these 5 points that are about as much as you need to know to be a succesful investor.
1. Save as much as you can so you can invest as much as you can.
2. Asset Allocation is king and should be based on one's willingness, ability, and need to take risk.
3. Avoid active management as it is a loser's game.
4. Realize costs, taxes, and inflation eat into long term returns.
5. STAY THE COURSE so you don't mess everything up.
As you will notice there is no mention of one valuation metric and that is the ironic part of succesful investing. You can be a GREAT investor without needing to know anything about stock valuations.
If you are interested in learning what counts in investing I would suggest start reading a couple simple books, either Allen Roth's "How a Second Grader Beat Wall Street" or Jack Bogle's edition to the Little Book series on index funds.
Thanks! I was told exactly this when being referred here. Level-headed, honest feedback. Presenting a contrarian approach is rarely received well in many forums that (as you mentioned) have their biases, but the quality of the feedback is really dependent on the members' ability to think clearly while writing out their responses. I appreciate that effort.
I'm pretty much on board with all of the 5 items you listed already, so I'm off to a good start! Maxed out employer match, maxed out ROTH IRA, and continue to save additional free funds monthly. Never keep more than is needed to be liquid in my bank accounts and the rest is invested in the market.
Asset Allocation is something that I'll need to work on. I have a significantly higher tolerance for risk than the average person, but its within my comfort zone and I don't lose sleep over it, so I'm not concerned there. Right now I'm divided up 8% healthcare, 28% financial industries, 16% consumer goods, 4% technology, 24% services, and 20% basic materials. Yes, it all in equities, but there is still some diversification.
The "active trading" part is interesting based on how you define it. Depending on your portfolio size, the fees associated with owning whole markets can exceed that of someone who executes, say, 100 trades in a year. This strategy employs 50 trades a year (25 buys, 25 sells), but that will total up to $400 or less on most online brokerages. For a $100k account, that's a 0.4% fee which may be comparable or even less to many funds. Obviously our goal is to keep as much of the profit in our pockets as possible, so this line will blur and eventually fall onto the side of the trader as account size increases. If you had a $10M account, for example, and could earn an 8% return through buying and selling 100 times per year or match the market's 8% return for a .15% fee, you'd be lighting over $14,000 on fire each year through the later method.
midareff wrote:Don't be offended and you obviously did a significant amount of work to get to the data you posted. As a group we are not interested in what stocks are the flavor of the month, or what latest super strategy produced XYZ results when back tested to 1491. As noted in prior posts we believe in saving as much as you can, asset allocation, owning whole markets, rebalancing, staying the course and avoiding active trading and other costs which reduce overall returns, among others as a brief list. There is much information on this site. Might be worth your time to delve into it, especially the suggested reading material lists.
Thanks! I certainly will. You guys seem like an intelligent bunch (though I sense at least a bit of group think, but that's to be expected from any forum given a long enough time frame), so I'll certainly stick around and learn and evaluate other methods!