Forums  > Basics  > Launching a less than ideal strategy  
Page 1 of 1
Display using:  


Total Posts: 250
Joined: Oct 2004
Posted: 2017-10-05 16:11
What are the funding options for “quirky” strategies? By that I mean, ones that give good annual returns, but have some attribute that makes them unappealing to most institutional investors. Let’s say, a large drawdown at some point, or a sharpe that is below 1.

Is it just personal and friends and family money? I have something I want to put in production, run for 3-5 years and then try to get some outside money, but not really sure who the market would be.


Total Posts: 478
Joined: Apr 2005
Posted: 2017-10-05 16:55
Sharpe sub 1 - people can live with, if correlation profile is interesting
Large DDs, less so


Total Posts: 250
Joined: Oct 2004
Posted: 2017-10-05 18:00
To be clear, this is all out of backtesting, but correlation with Sp500 is less than 0.1. Drawdown was 40ish percent, with average return to 18% annually from 1999


Total Posts: 1059
Joined: Nov 2004
Posted: 2017-10-05 18:01
jaiman, do you know that historical low max DD and high sharpe ratios are not predictors of future performance? If you meet the right investors, they will be more interested in what you do and how you do it rather than doing magic with your historical returns. So do not worry too much about these. I would be very concerned with a manager bragging about his low pas drawdowns. This guy would just be a disaster waiting to happen.

Let it run. Anyway your first point of failure is probably yourself. You may abandon it long before thinking of marketing it to anybody else.

If you are not living on the edge you are taking up too much space.


Total Posts: 250
Joined: Oct 2004
Posted: 2017-10-06 15:47
Ok, thanks Goldorak. Guess I'll get this thing going in January. Here's some stats, hopefully production is close!


Total Posts: 103
Joined: Sep 2015
Posted: 2017-11-13 21:58
Only knee jerks:

1 - is this the right benchmark? Your returns do not "look" like S&P 500.

2 - idk anything about your strategy - but can you couple it with anything else that reduces dd (even for your own sanity)?

I'd be curious to know your coskew to stocks despite your lowish rho/r2

Forum Captain

Total Posts: 1494
Joined: Jun 2004
Posted: 2017-12-21 21:17
@ contango

Could you define what exactly you mean by coskew in this context?

@ jaiman

You don't seem to make a distinction between in- and out of sample results. Why is that?

@ goldorak

I thought the point of out of sample results testing is exactly to establish relevance of backtesting for future results.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken


Total Posts: 459
Joined: Jul 2008
Posted: 2018-05-03 23:13
I think jaiman is right not to separate out of sample results. It's up to the person evaluating the strategy to provide the test data. Any "out of sample" period chosen by the person who came up with the strategy is suspect - you can't tell if jaiman looked at the "out of sample" data or not, and how many times.

BTW: reusable holdout
Previous Thread :: Next Thread 
Page 1 of 1