Backtesting Timing Strategies

or Register to post new content in the forum

19 RepliesJump to last post

 

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Dec 3, 2009 4:37 pm

Is there an easy tool or site you could use to backtest a timing strategy?  For example, if you wanted to backtest a moving average strategy, obviously you can get plenty of charts to identify buy/sell indicators.  But I can't figure out a way to measure the return of something like this over time.  Any ideas?

Dec 3, 2009 4:42 pm

Pen and paper

Dec 3, 2009 4:43 pm
B24:

Is there an easy tool or site you could use to backtest a timing strategy? For example, if you wanted to backtest a moving average strategy, obviously you can get plenty of charts to identify buy/sell indicators. But I can't figure out a way to measure the return of something like this over time. Any ideas?





You could download the data into either JMP or WINKS and backtest it. Or even Excel or Numbers.

Dec 3, 2009 5:03 pm

Yeah, I know the long way.  Just hoping for a shortcut.  Never used JMP or WINKS.  What are those exactly, and I assume they aren't free?

Dec 3, 2009 5:37 pm

What's the point of backtesting a strategy?

Dec 3, 2009 7:53 pm
anonymous:

What's the point of backtesting a strategy?



Historical returns, historical volatility, ....

Dec 3, 2009 8:03 pm
anonymous:

What's the point of backtesting a strategy?


 
OK, I'm sure you're just being sarcastic to prove a point, but the point is to determine how this particular strategy holds up under certain conditions.  Yes, historical results do not guarantee future results, blah, blah, blah.  I do not expect to experience the same returns, but patterns and trends can emerge.
 
Theoretically, unless you are just using annuities or some other fixed product, you MUST have backtested a strategy to come up with your allocations at some point, no matter hos simple the strategy.  How do you decide what funds or ETF's to use?  Ever look at historical results?  Never?  Really?
How do you decide which managers to use?  Which stocks to purchase?  Never looked at a stock's financial statements?  Yes?  You're backtesting THEIR strategy.
Dec 3, 2009 8:08 pm
B24:

Yeah, I know the long way. Just hoping for a shortcut. Never used JMP or WINKS. What are those exactly, and I assume they aren't free?





JMP and WINKS are statistical software. Great tools. I use JMP. Not free. Not even close. JMP is AWESOME for backtesting.



Take a class at the local university (if there is one) and you should be able to use one or the other for free.

Dec 4, 2009 2:43 am

Do you really talk to clients about stuff like historical returns, in today's world?

I'd rather talk to them about protecting them from a potential Biofreeze. Some idiot in the Middle East lets off a nuke, and every dairy in North America is shut down for a couple of years. Yeah, we still need to own cash, stocks and bonds, but are you joking about "backtesting"?

Dec 4, 2009 10:38 am
BioFreeze:
B24:
anonymous:

What's the point of backtesting a strategy?


 
OK, I'm sure you're just being sarcastic to prove a point, but the point is to determine how this particular strategy holds up under certain conditions.  Yes, historical results do not guarantee future results, blah, blah, blah.  I do not expect to experience the same returns, but patterns and trends can emerge.
 
Theoretically, unless you are just using annuities or some other fixed product, you MUST have backtested a strategy to come up with your allocations at some point, no matter hos simple the strategy.  How do you decide what funds or ETF's to use?  Ever look at historical results?  Never?  Really?
How do you decide which managers to use?  Which stocks to purchase?  Never looked at a stock's financial statements?  Yes?  You're backtesting THEIR strategy.




Loser.

 
Glad we could have this talk.
Dec 4, 2009 10:59 am
Milyunair:

Do you really talk to clients about stuff like historical returns, in today's world?

I'd rather talk to them about protecting them from a potential Biofreeze. Some idiot in the Middle East lets off a nuke, and every dairy in North America is shut down for a couple of years. Yeah, we still need to own cash, stocks and bonds, but are you joking about "backtesting"?



He's not talking about talking to clients about historical returns.  As everybody moves forward to GIPS standards, backtesting will become important for anybody who works in the securities industry. 

Statistics can help immensely with what B is talking about.  I can't understand why nobody wouldn't want to do it.

(Exception:  BioFreeze's business model has no need of it.  But the insurance companies that back the annuities will surely use backtesting methods to determine fees).

Dec 4, 2009 11:00 am
anonymous:

What's the point of backtesting a strategy?

 
Jim Simons said the following when asked if he tested his ideas first:
 
"Everything's tested in historical markets. The past is a pretty good predictor of the future. It's not perfect. But human beings drive markets, and human beings don't change their stripes overnight. So to the extent that one can understand the past, there's a good likelihood you'll have some insisght into the future. It's worked for us.."
 
That is from a guy who has averaged 35% annual returns since 1989 with only one down year... And that is after his fee of 5% upfront and 36% of profits...
Dec 4, 2009 1:45 pm

http://www.gipsstandards.org/about/governance/meetings/2009/sept/pdf/tab_17_usipc_suggested_practices_theoretical.pdf

For example, are you talking about trying to come up with a ( Morningstar Workstation) hypothetical illustration where no data exists?

Like, newer ETFs?

What do you do with recent comments from Bill Gross at PIMCO, who says we'll likely experience lower than average historical returns in the coming years?

Dec 4, 2009 2:00 pm

Remeber Bill Gross manages a bond fund and that everyone has alterior motives.

Dec 4, 2009 2:10 pm
Milyunair:

http://www.gipsstandards.org/about/governance/meetings/2009/sept/pdf/tab_17_usipc_suggested_practices_theoretical.pdf

For example, are you talking about trying to come up with a ( Morningstar Workstation) hypothetical illustration where no data exists?

Like, newer ETFs?

What do you do with recent comments from Bill Gross at PIMCO, who says we'll likely experience lower than average historical returns in the coming years?

 
Again, not trying to use this to present to clients.  It's not even to suggest certain results.  It's for me to better understand certain strategies and how they behave under certain market conditions.
 
 
Dec 4, 2009 2:12 pm
chief123:
anonymous:

What's the point of backtesting a strategy?

 
Jim Simons said the following when asked if he tested his ideas first:
 
"Everything's tested in historical markets. The past is a pretty good predictor of the future. It's not perfect. But human beings drive markets, and human beings don't change their stripes overnight. So to the extent that one can understand the past, there's a good likelihood you'll have some insisght into the future. It's worked for us.."
 
That is from a guy who has averaged 35% annual returns since 1989 with only one down year... And that is after his fee of 5% upfront and 36% of profits...
 
I'm hoping my strategy will allow me to charge this.
Dec 4, 2009 4:17 pm

Yeah that guy is insane.. hires no one from wall street, phds only(math, physics)...Check out his 13F he owns tons of securities.

Dec 4, 2009 5:35 pm

I confess, I don't even remember backtesting from the CFP courses, thanks for bringing it up.

For sure, ulterior motives abound, I'm thankful this career is part science and part art.

Dec 5, 2009 9:31 am

That would be true if only it were true.  I have not used an A-share for under 100K breakpoint since my first 6 months in the biz.  Nobody in my office pays more than 3.5% for A shares.  And I don't just use AF.  In fact I use them less and less these days.

 
But I know you were kidding.