It is attractive as a marketing tool for a portfolio that benefits from equity type appreciation while mitigating some of the downward or negative periods, which are temporary in nature, but very real to an investor that may have to withdraw periodically. As such is it suitable to market as a mutual fund alternative to relatively unsophisticated long term investors? Yes, probably. As you eluded to, yes, it is dubious in nature to a more knowledgeable individual, but that should not matter for the former.Dear colleagues! What do you think about this portfolio of Long/Short positions in the DAX universe (Cumulated returns of spread)? Average annual return: 20%. Maximum historical drawdown is 13%. Stationarity optimization + Fractal filter + CAPM optimization. Is it attractive for possible investment? Thank you.
Long/Short portfolio in DAX universe: stationarity filters
- Thread starter Sergey Kamenshchikov
- Start date
Thanks for your comment. Our #1 goal is marketing attractiveness of this instrument for passive unsophisticated investors. Thanks for comparison with mutual funds. "it is dubious in nature to a more knowledgeable individual". You mean that knowledgeable individual may need to know more about principles and algotithm of portfolio?
What's the portfolio turnover rate? [bump]
Allow me to elaborate: What we are speaking of here is essentially a claim of "alpha" generation e.g. the risk-free kicker/juice. I say this because what is clearly displayed is the "stability" (lack of draw-downs) relative to generic portfolio (index), of the strategy relative to its own performance.
...So, as a sophisticated investor might approach this: 1) what is the method of generating this alpha and is it realistically believable that whatever that method is might work not only in the recent past but into the future (?), and 2) in the absence of this, "are you selling me a golden goose" (?). I would expect a professional investor to have this approach and to not move unless given a clear answer to (1). But for mom & pop, presentation looks brilliant, just missing a "past performance" disclaimer.
Just as an observation, could be wrong here, but it seems to me looking at that curve you have one "filter" in particular that served to avoid the major drawdown(s) (a get out filter) & it his largely responsible for (probably) just about all your excess returns, I am guessing, and I'm also gonna step out on a limb here and say that filter is cherry picked / data-fitted.Dear colleagues! What do you think about this portfolio of Long/Short positions in the DAX universe (Cumulated returns of spread)? Average annual return: 20%. Maximum historical drawdown is 13%. Stationarity optimization + Fractal filter + CAPM optimization. Is it attractive for possible investment? Thank you.
Allow me to elaborate: What we are speaking of here is essentially a claim of "alpha" generation e.g. the risk-free kicker/juice. I say this because what is clearly displayed is the "stability" (lack of draw-downs) relative to generic portfolio (index), of the strategy relative to its own performance.
...So, as a sophisticated investor might approach this: 1) what is the method of generating this alpha and is it realistically believable that whatever that method is might work not only in the recent past but into the future (?), and 2) in the absence of this, "are you selling me a golden goose" (?). I would expect a professional investor to have this approach and to not move unless given a clear answer to (1). But for mom & pop, presentation looks brilliant, just missing a "past performance" disclaimer.
Thank you so much for this comment. Very fruitful ideas. 1) "are you selling me a golden goose". Totally agree. However I will be able to give only general ideas of optimization, because algorithm itself is novel and far from traditional (protected). I think about forward test: optimization through 2009-2014 and 2015 as observation peiod. It may help to analyze stability. What do you think? 2.) Past perfomance disclaimer will be provided
P.S. Thank you for the model. However I will need time to analyze it. And will ask you some questions from time to time. If you won't mind.
Portfolio is based on BUY&HOLD approach (no get outs). Yes, you are quite right. There are several filters combined with traditional optimization of weights. These filters provide stability and allow to remove portfolios with unstable behaviour from selection cycle. However it has 6 degrees of freedom (6 weights) only. Weights are provided to certain clusters of assets.
Hi! It's a simulation of BUY&HOLD strategy. We invested $100K last month just for analysis of real brokerage costs. However we will start the project and raising funds in December.
I'm confused, because original post said "Long/Short positions in the DAX universe"... Are you saying you your backtest bought a subset of the DAX in 2009 and held that portfolio ever since?
Sorry Not BUY&HOLD, buy ENTER/HOLD strictly speaking. Yes, it is LONG/SHORT basket.
Not BUY&HOLD, buy ENTER/HOLD strictly speaking. Yes, it is LONG/SHORT basket.