In the interest of continuous improvement, I took advantage of the summer break to think about how to further improve the profitability of the determining portfolio compared to the risks incurred. I will begin to introduce the first modifications in August, which I will provide below. Another should follow in September.
Real estate
Real estate has so far corresponded to a buy & hold position and was therefore not managed according to the principles of tactical asset allocation. This exception was due to the fact that the profitability of this type of investment is relatively stable over time. My various backtests with moving average did not manage to obtain better results than in b&h.
However, observing the variation of real estate prices in relation to the stock market, an idea came to me. Between these two asset classes, there is a strong correlation (0.74). Why not then resort, as for stocks in developed countries, to the joint use of a moving average and the unemployment rate, as explained in my book ?
After backtesting, not only do we get a slight outperformance, but we also lower the volatility, which is very useful, especially in the withdrawal phase.
Although in buy & hold, SRFCHA offers better profitability compared to the risks incurred than VNQ, the latter behaves better in tactical asset allocation. I will therefore focus on VNQ in the future.
Of course, if you prefer, you can continue to keep the real estate position in buy/hold, as was the case until now.
Nasdaq and other assets
By the way, I also took the opportunity to backtest the Nasdaq with the same strategy and, unsurprisingly, we get better results than with a moving average alone (this is not surprising since it is very strongly correlated to the S&P 500, with 0.9). I will therefore also use this strategy with the Nasdaq in the future.
In order to clarify the reading of the determining portfolio, I will now detail more precisely
- THE momentum from each position:
Average performance at 1, 3, 6 and 12 months
Price above the moving average, indicating positive absolute momentum
Price below the moving average, indicating negative absolute momentum
In green, rank of those who have the best relative momentum and who are retained
In red, rank of those who have the worst relative momentum and who are excluded
- THE unemployment American for each position:
OK, US unemployment trend favorable for this asset, no imminent danger
Asset that is not affected by the US unemployment rate
Warning, US unemployment trend unfavorable for this asset, indicating imminent danger
Cryptos
The year 2022 has been quite turbulent for the currencies adored by geeks. The determining portfolio has done well, because the positions (already minority) were sold between January and February. I still believe in the long term a return to grace of some of them, especially since we hear more and more about the collapse of cryptocurrencies and even sometimes their probable end. When the media cries wolf, in terms of investment, it is because the upward trend is approaching. Buy at the sound of the cannon...
That being said, I also believe in a consolidation of the sector and I will therefore focus in the future on the two majors: Bitcoin and Ethereum. I will therefore leave aside Solana, which I have also been following until now.
We are now well-equipped to tackle the second half of this year, whose autumn promises to be, as is often the case, particularly hot. At least in a figurative sense. For the rest, it will depend on the climate on one side and on the gas tap on the other.
Discover more from dividendes
Subscribe to get the latest posts sent to your email.
Français 
English 
Italiano 
Deutsch 
Español 
Português
Hi Jerome,
I'm not sure I understand this sentence:
"reduce SRFCHA from 15% to 6% of the value of the PF; it is also possible, in order to align directly with the PF, to sell all of SRFCHA and strengthen VNQ to 8% (this however generates additional costs)"
Is it in the sense that the current target is 8% so 6 + 2% is OK? And we will see later so that only VNQ remains?
I feel like the fees you're talking about will be paid sooner or later anyway if the goal is to move to VNQ alone.
On the other hand, would it not be wise to maintain the US + CH diversification?
THANKS.
Hi Thierry,
"Is it in the sense that the current target is 8% so 6 + 2% is OK? And we'll see later so that only VNQ remains?" Yes, that's exactly it.
"I have the impression that the fees you are talking about will be paid sooner or later anyway if the goal is to switch to VNQ alone." In terms of number of transactions, it is indeed the same, but in terms of volume the variant of fully settling SRFCHA and buying back VNQ is more significant, hence the slightly higher fees.
"On the other hand, wouldn't it be wise to keep the US + CH diversification?" In buy&hold yes, the two complement each other very well indeed. I would avoid using both in asset allocation, however, because SRFCHA does not respond well to this strategy. We could imagine keeping a share of SRFCHA in buy & hold and VNQ in asset allocation. However, my backtests give me better results with VNQ alone (in asset allocation). As mentioned in the article, those who prefer to remain b&h (either on SRFCHA or on both ETFs), can do so of course, it remains a completely acceptable approach.
Hi Jerome and thanks for your reply.
"In terms of the number of transactions, it is indeed the same, but in terms of volume, the variant consisting of fully settling SRFCHA and buying back VNQ is more important"
Hmm, so you anticipate that the share allocated to VNQ will decrease further or even reach 0%?
Because if the "final" state is >= 8% VNQ, there is something that escapes me in your reasoning.
Example with a 1000k portfolio:
Initial state: 15% S = 150k + 2% V = 20k
Final state: 0% S + 8% V = 80k
In 1 step: 2 transactions of resp. -150k and + 60k = 210k absolute
In 2 steps (or more):
A: 6% S = 60k + 2% V, so 1 transaction -90k
B: 2 transactions of resp. -60k and +60k
Total: 3 transactions: -90k -60k +60k = 210k absolute
So the same volume but at least 1 more transaction. So a priori more costs. If the allocation on VNQ never reaches 0%, it will be necessary to carry out 2 transactions at some point, unless of course you keep a small part of SRFCHA not included in the asset allocation.
"We could imagine keeping a part of SRFCHA in buy & hold and VNQ in asset allocation. However, my backtests give me better results with VNQ alone (in asset allocation)."
Thank you for this clarification.
The backtests you are talking about are VNQ alone in asset allocation vs VNQ & SRFCHA also in AA, correct? Or did you also test SRFCHA in B&H with VNQ in AA?
Thanks again and have a nice day.
Hi Thierry,
As the real estate position goes into asset allocation, and is therefore no longer in b&h, inevitably at some point it will also end up at zero. So if today you sell off all of SRFCHA, to buy back VNQ, then sell it all again later, it will inevitably give you more volume than if you only lower SRFCHA a little to match the target allocation for real estate. There are no "steps" to match the current target allocation, this will happen naturally when the allocation goes to cash at some point. So: 1 single transaction now to lower SRFCHA by 9 points, then, when the position goes to cash, 2 transactions (1 for VNQ by 2 pts and 1 for SRFCHA by 6 pts). In total, in the long term, 3 transactions for 17 points). If you change everything immediately, that gives you 2 transactions today (1 sale of SRFCHA for 15 points and 1 purchase of VNQ for 6 points), then in the long term, 1 sale of VNQ for 8 points. In total in this case, 3 transactions for 29 points this time. The difference of 12 points comes from the fact that there is a real estate portion of 6 points that is “unnecessarily” sold (SRFCHA) then bought back (VNQ) with the second variant.
VNQ in asset allocation beats, according to my backtests, 1)SRFCHA in b&h, 2)SRFCHA in AA, 3)VNQ&SRFCHA in AA, 4)SRCHA in b&h and VNQ in AA
“VNQ in asset allocation beats, according to my backtests, 1)SRFCHA in b&h, 2)SRFCHA in AA, 3)VNQ&SRFCHA in AA, 4)SRCHA in b&h and VNQ in AA”
Great, thanks a lot for these details, it's interesting!
Regarding the fees, OK I see. I had in mind that even in AA real estate could remain very high for a very long time. That said, after having done the exercise I arrive at the same conclusion as you even if it never quite reached 0%.
From a volatility and valuation perspective, there is little risk that the weighting will drop significantly. From a moving average perspective (& unemployment rate), there is also little risk that it will go to cash before a while. On the other hand, from a relative momentum perspective (best performing positions), even if real estate is doing well for the moment compared to other assets, it is quite possible that in a few months it will lose priority.
Hello Jerome,
Thank you for these explanations.
I can't trade VNQ. The possible permutations according to the user manual are:
VNQ: IUSP, SRFCHA (CH), REET (World), IPRP (Europe)
SRFCHA is a possible permutation of VNQ. Should SRFCHA be reduced to strengthen IUSP which is also one of the permutations?
I am still in the "gradual entry" period in the determining portfolio. So I have cash, do you advise me to keep SRFCHA, and in addition to strengthen VNQ (or IUSP in my case)?
Hello steph,
Yes, I advise you for now to keep SRFCHA and strengthen IUSP. The complete reallocation to VNQ (or IUSP for you) will be done later.
THANKS!