Comments on: Backtesting Investment Strategies

By: Monique

Monique — Fri, 02 Apr 2010 23:06:26 +0000

The idea of rebalance during election years is interesting and sounds fun, but not really anything base an investing decision.

By: WealthWebGuru

WealthWebGuru — Mon, 29 Mar 2010 18:24:30 +0000

One more item for your list of essential simple concepts is rebalancing. Buy and hold is the most common management strategy for an investment portfolio. Investors need to think about a strategy to manage a portfolio other than just buy and hold because buy and hold does not always work.

I think Otar’s book is excellent and is a wonderful resource for retirement.

By: Bret @ Hope to Prosper

Bret @ Hope to Prosper — Mon, 29 Mar 2010 17:32:55 +0000

I think back testing is extremely useful, if you are trying to sell a book on stock market investing. As for actual investing, I wouldn’t count on the past performance being an indicator of future results.

Without getting into the whole “efficiencies of markets” theory, here is the problem. Investors will quickly rush in to adopt any strategy that has been proven to be successful, in the past. This reduces the chances that any historic strategy will work similarly, in the future. Not to mention, the market conditions will change with time.

So, I use back testing for strategizing, but I never count on it.

By: Mike

Mike — Thu, 25 Mar 2010 16:46:22 +0000

His “buckets method” explanation starts on page 163. I’ll probably write a post about it in the near future, as I think it does a better job of minimizing the negative impact of volatility than lots of other retirement asset allocation methods.

“Just think what he’d say if he were from Belgium rather than Canada . . .”

Ha!

By: Larry

Larry — Thu, 25 Mar 2010 16:04:18 +0000

Mike: “I think his explanation of the significance of sequence of returns risk makes the book worth reading. And his take on the buckets method of asset allocation in retirement is quite good.”

Can you please expand on this? I must have missed it somewhere around page 463.

Mike: “Half as much allocated to Canada as to the U.S.?”

Just think what he’d say if he were from Belgium rather than Canada . . .

By: Mike

Mike — Thu, 25 Mar 2010 15:38:31 +0000

Indeed. I meant what I said about “important insights.”

I think his explanation of the significance of sequence of returns risk makes the book worth reading. And his take on the buckets method of asset allocation in retirement is quite good.

The book certainly got me thinking.

At the same time, the thoroughly bizarre scenarios he puts forth in his examples often made me wonder whether he was “cherry picking” in some way or other to make his points appear stronger.

6% withdrawal rate?
Half as much allocated to Canada as to the U.S.?
Assuming a 1% alpha over a 30-year period?

And he has a tendency to shift variables around from example to example, even when he isn’t making a point about those specific variables.

By: Larry

Larry — Thu, 25 Mar 2010 15:26:55 +0000

It should be noted that there’s a lot more to Otar’s book than the backtesting and re-balancing in presidential years, and in particular he has some provocative things to say about annuities (he’s for them) and asset allocation (he apparently thinks that doesn’t matter). He also takes 500+ pages to say what he probably could have said in 10.

By: Robert Wasilewski

Robert Wasilewski — Wed, 24 Mar 2010 19:47:48 +0000

Back testing has different uses. For example, anyone can come up with a profitable stock market system by using a screen that is continually tweaked until the outperformance is impresive. This of course is totally bogus and is why it is always important when viewing performance data to ask how it was achieved.
Using historical data to come up with appropriate withdrawal rates on retirees’ nest eggs is, I believe, useful. Since the 1920s the economy has experienced wild swings in inflation, stock performance, and bond performance. Withdrawal rates that worked well throughout should provide some comfort. And, more importantly, if they didn’t then they would be a source of concern.
Otar’s work does stress the importance of being careful in interpreting averages. Sometimes people get an unwarranted sense of well being because they are looking at averages. These can cover up the important impact of the market over the first few years of retirement. I’m sure anyone who retired at the end of 2007 knows exactly what I’m talking about.
Ask Long Term Capital Management about backtesting. That was how they came up with their so-called “value at risk” models.

By: The Biz of Life

The Biz of Life — Wed, 24 Mar 2010 17:33:31 +0000

Everyone seems to backtest to prove their investment strategy will work in the future. Of course, backtesting proves nothing other than giving the investor that wonderful feeling that they are using the perfect investment strategy for yesterday’s results. In the worst cases this leads to performance chasing. Just read the financial magazines in the grocery stores, classic performance chasing.