If understood correctly, they do weight their asset classes (commodities, bonds, forex) by some measure of risk parity. But within a basket everything is equally weighted. They can only go long or short. But they do so with trend-following against a moving average:
So the risk parity will not be able to positively influence picking long or short. Therefore this should only explain some volatility reduction and maybe a slight performance increase.
I went over the numbers once more and found my explanation. It could benefit from critically picking it apart. Thank you.
Bonds
I was wrong about the return on bonds. Their total return was double that of cash during the period from 1992 - 2023 (data from Portfoliovisualizer). Let’s calculate a rough factor by averaging the bond returns and dividing through cash returns. Like this:
factor = \frac{average(bond1, bond2)}{cash}
1992 - 2023
Asset
CAGR
Stdev
10-year Treasury
4.79%
7.29%
Global Bonds (Unhedged)
4.87%
7.52%
Cash
2.35%
0.59%
Factor
2.06
If we cut that up into 3 time frames:
1992 - 2000
Asset
CAGR
Stdev
10-year Treasury
6.62%
6.87%
Global Bonds (Unhedged)
7.37%
7.82%
Cash
4.58%
0.26%
Factor
1.53
2000 - 2008
Asset
CAGR
Stdev
10-year Treasury
7.14%
7.29%
Global Bonds (Unhedged)
8.25%
7.52%
Cash
3.26%
0.59%
Factor
2.36
2008 - 2023
Asset
CAGR
Stdev
10-year Treasury
2.66%
7.40%
Global Bonds (Unhedged)
1.90%
7.63%
Cash
0.75%
0.32%
Factor
3.04
Expected return
So let’s assume we only choose long and short according to expected return and don’t care to much about volatility. Also let’s assume handling cost of the underlying is negligible (would increase return of short commodities).
Commodities futures need to pay the risk free rate. The bonds need to pay that too, but had double the return of the risk free rate. Currencies need to pay it, but also return it.
Then our asset classes should return the following:
Asset
Position
Return
Weight
Leverage
Total
Commodities
short
1x
25.0%
3x
0.75x
Global Fixed Income
long
1x
42.5%
3x
1.275x
Currencies
-
0x
32.5%
3x
0x
Total
2x
Together with our collateral of 1x we get the expected 3x.
KFA MLM
The actual index follows another strategy. The above would probably have horrible volatility. Still it stands to reason that their elaborate strategy manages to reduce volatility. But in the end they don’t really manage to beat the fundamentals. That’s what I would expect from a strategy.
It would be interesting to know if the currencies actually contribute to the return (different from my assumption).
The coming market era might not be so kind to bonds. The worst case would be if they return the risk free rate. There would be no return on futures. If they perform significantly worse, going short will yield nicely. As all of this is coupled to the risk free rate, I would expect better returns, if interest keeps rising.
Some long-term data on 10-Y Treasury Bonds and 3-M Treasury Bills. Shown are the total return yields. Geometric averages centered on each year (looking forward and backward equally) over 1, 5 & 11 years.
Trend following / moving averages strategies seem to have a very good effect on reducing volatility for a rather small reduction in return (=> more leverage possible => more return).
If you speak German or can use a translation engine: Compare to ZahlGrafs Exzellente Abenteuer. In the last part of the series they also compare such strategies against vanilla portfolios. Still, more DD on my part is needed, before I believe it.
Concerning KMLM: I’m talking out of my ass. I found some data and read up on some concepts. They seem to give an explanation, but I just took the first explanation that fit the data. I’m by no means an accomplished hedge fund manager. But if my hypothesis is true, KMLM will not perform, because interest rates are going to rise and they are under-invested in commodities and over-invested in things that have no return. They did not manage to beat the fundamentals in 30 years. They will not in the future.
Do they have to beat the fundamentals, or do you think an uncorrelated stream of returns would be beneficial to your portfolio Sharpe in it’s own right?
I’m not comparing trend to equity, but to bonds, because I need to lower portfolio risk before FIRE (lower my equity beta to reduce sequence risk in retirement). Bonds can have like 40 years of negative real returns. Can trend too?
BTW, here’s some further food for thought. Let me know what you think of it if you got some time. It’s written by a trend shop, so beware TSMOM’s the buzzword, you might want to run a regression
Sure, but sharp doesn’t only have a divisor. If it dilutes the returns of the portfolio, because it doesn’t have any, your Sharp will not necessarily improve.
If the underlying premium disappears for an extended period, will trend still be able to deliver? Different doesn’t imply good.
Your paper says it has always delivered (with an annual 18% excess return before cost for a (monthly?) volatility of 10% since 1880). I will probably dig in deeper once I get some more time. If you find some more complete data, I’m interested. Their provided TSMOM data is only from 1985 and has only the mixed performance per asset class.
Also, thank you for finding this paper. I’m still interested to know, how and why did you find this information? I think there is so much useful knowledge out there, and it is so easy to not even notice that it exists. Maybe I can build some knowledge about finding things that I don’t know exist.
Excellent, looking forward to your thoughts on this. As to your question: I don’t work in finance, just very much interested in the field (also due to my early retirement plans). So I spend a lot of time (probably too much) reading financial research and other publications, listening to interesting podcasts etc. Everything publicly available online, no secret sources
That’s the thing. It’s all out there, free, possibly even condensed and well-written. But you will walk past it, even look right at, totally ignorant of what it is.
One shining example is this post on real estate by @oslasho. I read past it in April. A bit less taxes, how good could it be? In May I remembered (or reread?) the post and started digging. Lo and behold, the taxes are down to about 10% total on earnings (= cantonal tax + 1/2 federal tax, for legal entities). It is paid by the fund and losses are deducted. You pay nothing. No income tax on dividends, no wealth tax on shares, nothing.
But there is so much noise. Could be spam, things you already know, things that are useful for other people, things that are not enough useful, useful things that lack too much information or are too difficult for you. You can not follow up on all of them. And I wonder if this could be improved.
This forum still has an ok-ish signal per time invested for me.
How do you find the papers? I don’t assume you browse a database of papers for interesting sounding titles. So I guess the podcasts reference them then?
I recently found things by reading of potentially interesting words and concepts here, followed by digging with the search engine till I understand.
Maybe there even is good abstracted knowledge about this activity somewhere? Should be useful for an age where you don’t have information because it is buried by the flood, instead of it just not existing.
@Dr.PI Sorry, this is off-topic. Could you split our 2.5 posts into something like “Navigating the information flood”. I now have a concept, might start digging later. Others might have contributions, too.
I think we should not forget that bonds have had a historically unprecedented bull run of ~40years since the 80ies. So it comes at no surprise that trend returns can largely be explained by bond returns in backtests to the 90ies, since trend has been long bonds for that period.
Now what happens if this situation inverts, e.g. a 40 year bear market for bonds, hurting 60/40 investors big time? Trend, an opportunistic strategy, will move from long to short bonds and capture the returns.
Also, lets not forget that trend does not only consist of US treasuries, but diversified bonds, thereby reducing single-asset risk. And more importantly, apart from bonds it also holds very diverse commodities and currencies, with different trends to capture. So it’s not like just levering up a single asset.
This makes it far less path-dependent. Which is incredibly important for mitigating sequence risk in retirement. Bonds suck for 20 years? No problem, trend goes short. Bonds AND commodities suck for 20 years? No problem, trend goes short both. And then there’s also currencies in the mix. To me, that seems far superior to holding bonds & commodities long only.
There definitely seems to be a risk parity element to trend. But, in comparison to classic risk parity strategies, trend adds a second dimension by being able to go short assets. This could lead to more diversification compared to long-only risk parity strategies.
Indeed, expected returns are very hard to tell, if not impossible. We just don’t have enough data. But same holds true for equity.
I cite Robert Novy-Marx:
"When I talk to investment advisors, I love to ask them, “What is the equity premium?” I get a range of answers, but almost everyone thinks it’s somewhere between 4 and 7% and 6% is what I hear most. Then when you ask people where they come to that number, they say, “Well, you look at the last 100 years and it’s been 6%.”
That’s true. If you look at 100 years of data, you see a 6% equity premia, but if you get that mean equity premia by running the regression, it also gives you a confidence bound on your estimate. It turns out that the estimated equity premium that you get by looking at the past data is 6% plus or minus the standard deviation of like two and a quarter percent.
So, if you just are a frequentist probablist, you do the frequentist statistics, you’re basically saying you’re 95% sure that the true equity premium is between 1.5 and 10.5 percent. You just have no idea. With 100 years of data, we can’t come close to agreeing on what the equity premium is."
In my proper words: In finance, nobody knows nothing, and anyone who tells you otherwise is BSing. It’s not an exact science. Markets and prices are social constructs, purely man-made, do not adhere to natural laws like physics. So imo, the best and only bet we got is diversification.
Can anyone tell me if the capital gains distributions from funds like KMLM are subject to US witholdings taxes? And if so, are we able to claim them back?
Because they are considered capital gains mostly and are not taxed by Switzerland, I could imagine that we cant claim them back with DA-1?
As far as I know, nearly all US domiciled ETF get US L2TW taken from their distributions. There are certain exceptions like QII (Qualified Interest Income) discussed in this Bogleheads thread. They also link a iShares document that details for every fund how many percent were exempt for each distribution. From their wiki I gather, that the second exception are distributions from short-term capital gains.
But do you want to see a magic trick? I found that the DBMF ETF has spawned a mutual fund twin in Luxembourg. That means no L2TW and probably no Swiss taxes (maybe on collateral bond interest).
Sadly, if you are not already somwhat rich you can only buy their C-class shares with 1.60% management fees. Their I-class shares need a minimum investment of 1’000’000 USD. Edit: You can also buy the much better R-class shares (see below).
Just to see if it works at all, I placed an order on IBKR. Seemed to go through, as in the money is gone. We will know in a few days what I will get.
The somewhat better option considering the rather high ratio of distributions would be selling the day before and rebuying on the ex-dividend date. In the last few years DBMF and KMLM have only had a single distribution or less per year.
Some calculations with (short) historical data from Yahoo Finance give somewhat acceptable costs around 0.10% for DBMF and some huge gain at KMLM.
Calculation Table
ETF
Previous Day (Pre)
Ex-Date (Ex)
Minimum Volume
Close
Open
Average Pre
Average Ex
Distribution
Result Close- Open
Result Average
DBMF
2023-12-26
2023-12-27
241800
26.55
25.87
26.5725
25.835
0.75
-0.26%
-0.05%
DBMF
2022-12-27
2022-12-28
370000
31.19
29.07
31.3075
29.09
2.247
-0.41%
-0.09%
DBMF
2021-12-29
2021-12-30
30300
28.25
25.14
28.2975
25.6775
2.677
1.53%
-0.20%
DBMF
2020-12-28
2020-12-29
100
25.41
25.41
25.41
25.41
0.202
-0.79%
-0.79%
KMLM
2023
none
KMLM
2022-12-27
2022-12-28
115800
34.4
30.88
34.4025
30.6975
2.481
3.02%
3.56%
KMLM
2021-12-28
2021-12-29
300
28.78
26.87
28.78
26.88
1.838
0.25%
0.22%
Calculation Information
The “average” price has been calculated from the average of Open, High, Low, Close. It should represent a price you could get on that day.
Both results are calculated as follows:
result = \frac{(price_{pre} - price_{ex} - distribution)}{price_{pre.average}}
They are the percentage change for doing this transaction instead of receiving the distributions (before withholding tax).
I don’t know what the Swiss taxman will say about it though.
I heard that there is besaically no worthy european managed futures fund and 1.6% TER is insane and does not make it worth imo.
I also recently looked at wirtholding taxes and apparently the US does not withhold longterm cap gains distributions, which KMLM‘s distributions are. Cant cite it atm though. On ictax tgere is also nothing that says tax is withheld.
But I will this year sell almost all before ex-day and just keep a few share, then re-buy (or buy another fund like CTA) and will see what happens. Should be no problem if I do it once.
I saw that, but those should be relatively miniscule, if any at all. Last year for example there wasnt any distribution at all. The bulk will still come at year end, due to the nature of how futures work, as those cash settle then.
Same fund, but the “R” share class LU2572481948. I first thought “R” is for retirement funds and did not consider it (probably because of US mutual funds). It is not.
I placed an order for 257248194 IMGP - DBI MANAGED FUTURES “R” (USD) ACC on IBKR for 200 USD. It accepted. Maybe because IBKR uses an omnibus account? Results should be available by the end of this week.
The “R” class has only a maximum of 0.80% management fees. Redemption fees are 1.00% for all share classes.
By reading and partipating to this forum, you confirm you have read and agree with the disclaimer presented on http://www.mustachianpost.com/
En lisant et participant à ce forum, tu confirmes avoir lu et être d'accord avec l'avis de dégagement de responsabilité présenté sur http://www.mustachianpost.com/fr/
Durch das Lesen und die Teilnahme an diesem Forum bestätigst du, dass du den auf http://www.mustachianpost.com/de/ dargestellten Haftungsausschluss gelesen hast und damit einverstanden bist.
?
