Thanks!
I find the return stacking instruments quite interesting. How do you go about allocations? E.g when using trend following 30% allong VT (or similar). How do you look at this with these leveraged products?
1 Like
After all it‘s 100% stocks + 100% managed futures (at ~15% volatility, not every managed futures fund is created equal), and that‘s how I incoporate it into the portfolio.
Of course if you wanted 70% stocks and 30% alternatives, and you use RSST for that would them put you at 100% stocks and 30% alternatives, introducing significant leverage and bringing you to 100% stocks again.
You could however do 70% stocks of which 30% is RSST (or split that with Winton for example 50/50) and then use the other 30% for other uncorrelated strategies, maybe a multi strategy approach for example. And without the need to manage a margin loan and associated risks.
Then you are 70% stocks, 30% managed futures, 30% other alternatives.
If correlations are similar to history, this will increase return, without much increase in risk metrics compared to 70/30 alone, therefore higher Sharpe.
This is basically what I do with my portfolio.
Isn’t there also the Man Trend Alternative IN USD Acc, LU0503879685:USD summary - FT.com
AFAIK the market leader in trend follower.
That‘s an awful ucits version of their trend program. Extremely high fee with on top performance fee and low volatility (worst of all worlds essentially).
Performance therefore has been trash.
It‘s one of the older ucits offerings, when back then these funds were garbage here.
E: At least it has been in the past with the available share classes. Maybe if you can get your hands on the institutional share classes with lower fees, it can be worth it. I‘ll check once I have time on what is available. Maybe they updated their offerings since the last time I checked, due to the rising competition of better offerings. For a low enough total fee, the funs is interesting.
—————
MAN doesn‘t have good ucits offerings currently as far as I know. They have been market leader in that space in the past in the US. Not sure if they are still. But they are a good player (in the US). AHLT managed by them is a good etf for example.
The top providers are part of the SG trend index, the constituents you can view here: https://wholesale.banking.societegenerale.com/fileadmin/indices_feeds/SG_Trend_Index_Constituents.pdf
Worth zooming out from fund selection for a moment.
What is being discussed as “trend-following” is, in academic terms, time-series momentum (Moskowitz, Ooi & Pedersen, 2012) — the finding that the sign of an asset’s return over the past 1-12 months predicts next month’s return, and that this works across asset classes: equities, fixed income, commodities, currencies. Managed futures funds (CTAs) package this signal as a standalone product. That is, however, only one way to use it.
The alternative is to apply the same signal at portfolio level — using momentum and cross-asset correlation to drive allocation decisions across the whole portfolio (equities, fixed income, commodities, etc.) rather than running a separate trend-following bucket alongside an equity ETF. The two approaches compound differently:
-
Trend fund alongside equity ETF. The trend fund is a separate position, sized at e.g. 20-30%, intended to make money when equities fall. When trend works, it offsets drawdowns (2008, 2022). When it does not (most of 2010-2021, much of 2023), it drags on the portfolio. This is the source of PhilMongoose’s whipsaw concern — and Patirou’s intuition that you need a long psychological tolerance for the years when nothing happens.
-
Momentum applied to the whole portfolio. Instead of a fixed equity-plus-trend split, the momentum signal decides which asset classes to hold and at what weight every month. Drawdown protection comes from the portfolio adapting (cutting weak asset classes, holding stronger ones) rather than from a counter-cyclical position. No separate trend bucket; the signal is built into the asset allocation itself.
Neither is unconditionally better — they answer different questions. If the existing portfolio is 100% equity and the goal is crisis insurance, adding a trend fund makes sense. If rebuilding from scratch with a multi-asset framework, applying the signal to the whole portfolio is more capital-efficient.
On Patirou’s “weed out too-good-to-be-true funds” question — the useful number is not the long-run correlation to equities, which can be near zero in calm periods and very different under stress. Look at how the fund actually behaved during specific equity drawdowns: Q4 2008, March 2020, 2022. If a fund only looks uncorrelated when nothing is happening, that is not the protection you are paying for. AQMIX and DBMF have public histories long enough to check this properly. Winton Trend Enhanced Equities is structurally interesting precisely because the trade-off between equity exposure and trend exposure is explicit, rather than a passive blend.
Two honest caveats:
-
Trend strategies suffer in sharp reversals — the documented “momentum crash” problem (e.g. April 2009, late 2020). A dedicated trend fund concentrates this risk in exactly the part of the portfolio meant to be the hedge, which is awkward.
-
Backtests of trend funds carry the usual issues plus regime sensitivity. A single decade can look heroic or terrible depending which decade you pick. The 1970s-1990s data the original German articles lean on includes a structurally different interest-rate environment.
Worth doing the empirical work on whatever specific allocation is being considered, rather than picking 20% or 30% by feel.
-–!
Use of em dash suggests AI, but it doesn’t look like slop!
Looks very much like it. Maybe this shows my limited view of humans, but it’s hard for me to believe that large parts of this post (as well as the other post by the same user) has been written by a human.
1 Like