In April 2022, I introduced on this blog The Italian Leather Sofa Model Portfolio. As a reminder, here is the portfolio composition:
- 60% stocks (via NTSX)
- 40% bonds (via NTSX)
- 20% trend (via DBMF)
- 10% commodities trend (via COM)
- 4% Tail risk (via TAIL)
- -34% cash
The idea behind the portfolio is stolen from here. The link offers the best explanation of what I think is the most common question related to it, i.e. why the portfolio uses leverage (and why, in this context, leverage decreases risk).
It represents a simplified version of the portfolio I have been building since I moved to Switzerland: here you can find details about the “enhancements” to this model.
Please note that the returns you find in the Model Portfolio series will always reflect the point of view of a USD-based investor. The ETFs are priced in USD and Testfol.io, the app I use to track the portfolio, does not allow me to change the reference currency.
Besides these ‘technicalities’, the focus of this series is on how to build a great and simple permanent portfolio. There are various solutions an investor can employ if they do not have the USD as their base currency and want to eliminate the FX volatility. As I wrote here about the All Weather Portfolio, I am not bothered by the FX risk, given my investment horizon and the fact that I do not consider myself a CHF-based investor even if I live in Zurich. Plus, I do not have any currency-specific audience that would make this series more helpful if run in EUR, CHF or GBP (if you want a deeper dive into FX risk, I wrote this).
After an 8.78% in Q2-26, the portfolio went back to a more normal (yes, sure) 1.32% in Q3-26:
Since inception plus backtest (May, 2019).VBAIX is the 60/40.
The blue line represents the Model Portfolio, while the other two are functional references (I cannot really call them benchmarks): the 60/40 portfolio (yellow line) and the S&P500 (red line).
Since inception plus backtest (May, 2019).VBAIX is the 60/40.
Q3
Below you can find details of each ETF performance, including dividends, in the quarter:
Here is the Q3 price graph for each component of the portfolio:
How to read the portfolio performance
I have to admit I fell for the single-line item performance fallacy. NTSX is the ETF with embedded leverage that allows the addition of “free diversifiers” to the portfolio. I, wrongly!, judged the merits (or otherwise) of leverage within NTSX, thinking, for example, about the implications of an inverted yield curve (NTSX borrows at the short-term rate and invests in bonds that pay the long-term rate…not great when the curve is inverted).
Leverage belongs to the portfolio.
Not only that. COM and DBMF use futures; a small fraction of the sum invested in those ETFs is posted as margin while all the balance erns the T-Bills returns. In other words, if the Bills rate is 5% and DBMF returns 3%, it means DBMF alpha, the real yield of the strategy, was -2% for that year.
Over the last few years, there has been plenty of debate about the death, resurrection and second death of the 60/40 portfolio. What has received far less attention, at least with the same intensity, is that a better alternative exists: stocks, bonds and trend following.
This is the new 60/40: NTSG and MFEH. There is no easy-to-find backtest data today, but the ingredients are pretty basic. Trend following is only a novelty for ETF investors, not for the industry, and the SG indexes have a 26-year track record. That is long enough that in some European countries you would qualify for a state pension. It’s like showing you pasta, tomato sauce and onions. Nothing is cooked yet, but you can already imagine the taste of the finished dish (I should add that I’m Italian, so you’ll get that kind of taste. Not every kitchen manages to turn three simple ingredients into something great, hello Germany).
So yes, there is no backtest for NTSG and MFEH yet, and still I can tell you that in five years people will “discover” them, just as they are doing today with NTSX and DBMF.
I still remember the pushback when COM was doing this:
“it is broken”, “it has been arbed away”, “it was just datamining”.
Opsie Daisy
When you face all-time highs, you can no longer push the narrative that the strategy is broken. You have to find a different issue, like what happened to CTA (the ETF). But that’s why the strategy works: it climbs the wall of worry. I’m not saying it’s impossible that the manager behind COM will lose his mind (or the one behind DBMF), but that’s a manageable risk, just as it was for CTA.
This morning I was listening to this interview with BigERN and….surprise! From the man who’s not even able to distinguish between momentum and trend (sarcasm /off), he offered a quite compelling idea on why the strategy is still working.
As much as we like to toy with numbers and data, ultimately you either see it or you don’t. There will never be an “ultimate smoking gun” that closes the debate, and that’s for (at least) three reasons.
First, there will always be a better-performing asset, which means there will always be people asking why they should bother with diversification at all. Second, many investors are blinded by one or more biases, and no amount of data will change that. Third, and most intractable, are those whose identity comes from a framework. I suspect BigERN falls into this camp. For them the problem isn’t evidence; it’s that breaking free would mean enduring a level of pain they will never be willing to accept.
I build and destroy theses myself: if the objection to trend following is really about evidence, why was BigERN so eager to embrace the recurring sale of 1DTE options as a diversifier? Ain’t that out of his character?
As I said, it’s probably not worth spending time asking why others don’t see it. I’m glad they don’t, in a way, because you always want smart people challenging your thesis.
Anyhow
The biggest risk today for the Model Portfolio is that Meb closes TAIL… just before the event it was built for happens. That’s the risk. But find me the pundit who is talking about it. No one is, because they’re all too busy looking at what happened yesterday instead of concentrating on what matters: being ready for tomorrow.
The biggest positive for the Model Portfolio is that bonds are loading up future returns, with yields moving higher and the curve steepening. The portfolio is making 12% a year while part of the engine is storing energy instead of sending it to the wheels, which is the whole point of a diversified portfolio.
A reader on our subreddit asked how it was possible that NTSG was shielded from the recent bond rout. It wasn’t, of course. But the question made me realize how strong the single line-item bias is. It also made me realize how well stacked products solve that problem, in a jiu-jitsu sort of way.
DBMF has launched its own stacked version, DBAR: 100% DBMF plus 30% SPY (or something close to it; I didn’t check). I didn’t check because I plan to move my personal DBMF allocation to MFEH shortly, depending on what the EUR/USD trend does. Do not try this at home! Maybe one day we’ll get a 100% MFEH plus 30% VT. One can only hope.
What I am reading now:
Follow me on Bluesky @nprotasoni.bsky.social
0 Comments