I have not done a lot of reading on any of that, so I can't comment. MKLM and DBMF seem to be popular among some experts and hobbyists, but I think the expenses are still borderline high. My understanding from my brief reading of some recognized expert on X (formerly Twitter) was that trend following algos or shall we say rules are more or less just simple time lag momentum (price difference) numbers averaged over a few different time lags between 1 and 12 months. A high school kid can implement it. I may be wrong, but I may also be right, because complexity rarely results in tangible benefits in finance. You could automize the data collection and trading with Interactive Broker's or Schwab's API.Thanks comeinvest. I have some followup questions if I may.Either use the above mentioned ETFs leveraged with box spreads, or implement trend and carry yourself with futures. The latter would require a good amount of studying, but should be doable. I think the underlying rules are quite simple for trend. Interest rate carry you already have for the most part, because STT not always but usually have the highest carry, as physicist demonstrated in a few post in this or related threads.I have read the whole thread and benefitted enormously from the work of the participants, have to name skier and comeinvest specifically (no disrespect to others). I have various questions, but perhaps the one that bugs me most is about incorporating trend following and other alternative strategies into mHFEA. There were a few questions and responses after the quoted posting above, but there is quite little discussion about trend following and managed futures so far....
Has anyone done quantitative studies about how to add trend to mHFEA and to what amount? I know it is hard to be exact, but let's say we have a 120% equity/80% ITT portfolio. How would I reshape this portfolio to include trend? Of course I should give you a metric to optimize the porfolio on. I think I am most afraid of the maximum drawdown, hence the "modest" 2x leverage, but sharpe and sortino are also important. This is probably still too vague, but any help is appreciated.
I would ask the same thing for "carry" or "futures yield". I think carry has various implementations as trend, but what I am interested in is the context the term is presented by Return Stacked based on the Bloomberg GSAM Cross Asset Carry Index: https://www.returnstacked.com/a-differe ... enchmarks/. This form of carry has low correlation to both trend and stocks, similarly to the low correlation of trend to both stocks and bonds. So, putting everything together possibly has high risk adjusted returns, which we can leverage on (or can we?). Though, I should mention that I have not seen the correlation table of all 4: stocks, bonds, trend, yield. One down side is that, I am not even sure how many yield ETFs are out there, other than the upcoming RSSY.
These "alternatives" look scary to me since they are forms of active management even when they are implemented in a rules based way. I am aware that reputable investors like AQR and Alpha Architect are very positive about trend (not sure about carry). Moreover, some detail oriented people here who discuss how costs of 20bp can be eliminated for hundreds of posts seem to have a favorable view as well. However, the fact that the implementations are various and different is a challenge. Even if I decide that these are good ways to diversify and quantify the amount to invest, which instruments do I use? LazyOvrthinker in this thread and others in the Rational Reminder forums seem to prefer a mix of the most reputable ETFs like MKML and DBMF, but can we trust these in the long term?
There is a lot to discuss, but I should probably leave the intrinsic value of these strategies to other dedicated threads. My main question right now is how trend, carry or similar "alts" can be parts of mHFEA.
I would say the expense ratios of MKML and DBMF are lower than traditional cost for those kind of strategies, but still borderline high. Expense ratios matter because the future benefit and risk are unknown although we hope the risk-adjusted return is positive, while the expense drag is certain.
1- Since you specifically mention these ETFs, I assume you think they are pretty decent. Do you have any other suggestions? There are about two dozen trend ETFs mentioned on Rational Reminder forums, but only a few are discussed in any detail.
2- About self-implementation, we can follow the general time series momentum principals for our own treasury or SOFR futures, yes, but I remember someone (likely LazyOverthinker) mentioning that that is not really what you want. The "magical" trend following gains that save you in a crash show up when you diversify on many different futures markets. I think that is what AQR and Alpha Architect documents say as well, though it has been a while since I read them. I guess the issue is the same with carry, the treasury and SOFR futures you have studied in such detail forms only a part of what constitutes the Bloomberg GSAM Cross Asset Carry Index. Of course, this just means I just go to the ETFs, but I wanted to make sure I am not missing a point about DIY trend following or carry.
3- How do we decide on how much of the portfolio goes to trend ETFs? We can backtest, but ETFs are not very old. There are some older indices and studies (numbers are not always easy to get). Would you just use their expected return and volatility to have an optimum mean-variance portfolio, hoping the ETFs would roughly capture the historical results? Anytime I start thinking about these implementation issues, it feels like I am leaning on wishful thinking rather than a solid plan. The details you go into in this thread increased my standards I guess.
As for the position sizing and asset allocation, like I said, do your research which will probably lead to fuzzy results, then settle on a strategy. Please post your backtests here.
Statistics: Posted by comeinvest — Thu Apr 04, 2024 1:43 am — Replies 3004 — Views 529067