Managed Futures Strategy ETF DBMF

[Always passive]

Question: does a small percent of Managed Futures Strategy, like DBMF, that has a negative correlation to equities and fixed income fit in a portfolio?

[nisiprius]

Question: does a small percent of Managed Futures Strategy, like DBMF, that has a negative correlation to equities and fixed income fit in a portfolio?

1) Nothing "has" a negative correlation to equities and fixed income. Some things "have had" such a correlation. This is important because of the instability and unpredictability of correlations.

2) You always have to answer the question, is a "small amount" of anything going to move the needle? You won't get big benefits from small commitments. To make a big difference either way you are going to need to make a big commitment and also take a real risk if you are wrong.

3) Low and negative correlation is not automatically beneficial. The conditions under which they are beneficial are fairly strict. Consider a portfolio P and a proposed diversifier D. D will only improve P if it has low correlation and roughly comparable return and roughly comparable standard deviation. It's not unusual for such a thing to persist for a few years, but unusual for it to persist for a long time.

4) This is a chart of Morningstar category averages for managed futures funds (they now call the category "systematic trend.") This is how the average managed futures fund has actually performed. To compare apples to apples, I compare it to the Morningstar category average for intermediate core bond funds.



Based on actual average of real funds, I don't think managed futures look very appealing, despite a magnficent recent run-up. Do you think 2021-2023 make up for 2008-2021?

(The typical response of managed futures fans in the forum has been to say "we aren't talking about averages, we're talking about this fund" which typically is one without a very long history...)

[Always passive]

Thank you. The graph tells it all

[Gecko10x]

The trouble with managed futures is that it isn't "one thing", so it's hard to generalize. Some MF funds will include stocks, so that will make them more correlated to the stock market, but may boost performance when nothing else is trending. Other MF funds will include/exclude currencies as well.

I'm a fan of MF because it's another potential driver of returns that usually is uncorrelated to stocks & bonds, so it has the potential to help when nothing else is, while not being too much of a drag on the portfolio (unlike volatility funds).

[Professoro]

Question: does a small percent of Managed Futures Strategy, like DBMF, that has a negative correlation to equities and fixed income fit in a portfolio?

1) Nothing "has" a negative correlation to equities and fixed income. Some things "have had" such a correlation. This is important because of the instability and unpredictability of correlations.

2) You always have to answer the question, is a "small amount" of anything going to move the needle? You won't get big benefits from small commitments. To make a big difference either way you are going to need to make a big commitment and also take a real risk if you are wrong.

3) Low and negative correlation is not automatically beneficial. The conditions under which they are beneficial are fairly strict. Consider a portfolio P and a proposed diversifier D. D will only improve P if it has low correlation and roughly comparable return and roughly comparable standard deviation. It's not unusual for such a thing to persist for a few years, but unusual for it to persist for a long time.

4) This is a chart of Morningstar category averages for managed futures funds (they now call the category "systematic trend.") This is how the average managed futures fund has actually performed. To compare apples to apples, I compare it to the Morningstar category average for intermediate core bond funds.



Based on actual average of real funds, I don't think managed futures look very appealing, despite a magnficent recent run-up. Do you think 2021-2003 make up for 2008-2021?

(The typical response of managed futures fans in the forum has been to say "we aren't talking about averages, we're talking about this fund" which typically is one without a very long history...)

Plenty of wrong information here from last month:
1. Managed futures almost always had negative correlation with equities, particularly during equity drawdowns. Correlations may change temporarily, but that does not change the fact that managed futures have negative correlations with equities. How much of the following chart is above 0?


2. Managed futures + equities beats bonds + equities both on returns and portfolio risk. The main reason is the negative correlation of managed futures to equities during drawdowns (crisis alpha).


3. Not one managed futures fund, but almost all of them do better than bonds. The one above is for KMLM which goes back to 1988. Here are the comparisons with AlphaSimplex and AQR managed futures funds from the last 10 years:

[Always passive]

Very informative, thank you.
On KLML, I only found data from 2021, when the ETF seems to have its start. Where did you find data from the late 1980s?

[Always passive]

Very informative, thank you.
On KLML, I only found data from 2021, when the ETF seems to have its start. Where did you find data from the late 1980s?

[nisiprius]

...3. Not one managed futures fund, but almost all of them do better than bonds.

The chart I posted is for Morningstar's category average for all managed futures fund. How can the category average underperform bonds if "almost all of them do better than bonds?"

[nisiprius]

Very informative, thank you.
On [KMLM], I only found data from 2021, when the ETF seems to have its start. Where did you find data from the late 1980s?

I have the same question. According to the provider, KFA Funds, the inception date for KMLM was 12/2/2020. Source

[Alpha4]

Very informative, thank you.
On [KMLM], I only found data from 2021, when the ETF seems to have its start. Where did you find data from the late 1980s?

I have the same question. According to the provider, KFA Funds, the inception date for KMLM was 12/2/2020. Source

The underlying index (the Mt. Lucas MLM Index) goes back to 1988...actually, there are even backtests of it to 1979.

[ScubaHogg]

...3. Not one managed futures fund, but almost all of them do better than bonds.

The chart I posted is for Morningstar's category average for all managed futures fund. How can the category average underperform bonds if "almost all of them do better than bonds?"

Well if a few absolutely get crushed in comparison to bonds it could drag the average down for all of them I guess?

[Professoro]

Very informative, thank you.
On [KMLM], I only found data from 2021, when the ETF seems to have its start. Where did you find data from the late 1980s?

I have the same question. According to the provider, KFA Funds, the inception date for KMLM was 12/2/2020. Source

1. KMLM is the etf that was launched. As Alpha4 mentioned MLM index goes back to 1988. The monthly return data is public in the second link. Anyone with a paid subscription to Portfoliovisualizer can create the fund.
https://www.mtlucas.com/mlm-index
https://engage.kraneshares.com/s/77b9d7 ... mlm&page=1

2. I don't know what Morningstar considered as managed futures in the past. It looks like in 2021, they renamed the managed futures category and assigned someone to it who knows what managed futures (trend following) is , Samuel Lee. His article on managed futures is well written and section 6 explains the role of managed futures in a portfolio: https://advisor.morningstar.com/uploade ... tegory.pdf
https://advisor.morningstar.com/Enterpr ... FAQsv2.pdf

3. The evidence on trend following goes back 100+ years. One can learn what it is and why it works from this article: https://www.aqr.com/Insights/Research/J ... -Investing

4. The index for trend following is the SG Trend index. Unfortunately, the data is not freely available anymore. Someone posted the returns for SG trend, SG CTA and the main actors in the trend/cta space on another forum. I am copying his table. It is not complete as some funds such as Dunn Capital only has data that is limited to MFTNX's history. They have been doing this for over 40+ years. Mulvaney has been around for a long time too.
https://wholesale.banking.societegenera ... s-indices/

https://community.rationalreminder.ca/t ... 15439/3454


https://www.trendfollowing.com/2022/09/ ... rformance/

[nisiprius]

Another relevant detail is that KMLM's expense ratio is 0.90%, and DBMF's is 0.85.

[Professoro]

Another relevant detail is that KMLM's expense ratio is 0.90%, and DBMF's is 0.85.

Correct. However, their approaches are significantly different. KMLM is a pure trend fund using future contracts on commodities, currencies, and bonds. DBMF uses a replication approach. Some funds such as AQR's QMHIX also uses trend on equities in addition to the three asset classes I listed. Some people use their own trend following but managing futures contracts is challenging for people that have regular jobs. Stacking the trend following with equities and bonds are also getting popular with the success of funds such as BLNDX. Other funds in this space include RSST, RSBT. These funds do what they advertise to do so one can decide if it belongs to their portfolio.

It is not a managed futures fund, but here is another fund that some would detest, BTAL. The inception of the etf is 2011 but I have historical data since 2001. It has lost money since its inception and its net expense is 1.43.


From a basket of funds including, total equity, total bond, small cap value, large cap growth, trend following, btal, here is the optimized portfolio for max returns at 7% annual volatility since 2001.







A 60-40 portfolio is not diversified, neither the lifecycle funds that most people put their money retirement money in. Even during a time period with lowering interest rates, bonds did not provide protection during significant drawdowns.

[Professoro]

Excellent post on the diversification and lost decades risk: https://www.reddit.com/r/LETFs/comments ... k_setback/
VFINX is used for "Stocks".
VUSTX is used for "Bonds".
^GOLD is used for "Gold".
KFA MLM index is used for "MF".


Here is another comparison of stocks+trend vs stocks+bonds since 1985. I was able to construct it using Dunn Capital's returns which goes back to 1985 compared to MLM which goes back to 1988. Dunn has a higher volatility compared to MLM. Note the worst year and max drawdowns.



https://www.autumngold.com/capital_trad ... ack_record

[pathikrit]

FWIW, here (https://testfol.io/?d=eJy9j0FLw0AQhf%2F ... fgLC%2F6Xg) is a 30-year backtest with 15% TQQQ + 75% KMLM + 10% UGL.

[mbouck]

Markowitz was right. "The only free lunch in investing is diversification". Why this works, at a simplistic level, is multiple uncorrelated return streams. This is how AQR, et al. work their hedge fund magic. Well, that and leverage. Basically they take a *lot* (we're talking 20+) of uncorrelated return streams (e.g. trend, long-short, macro/options trading, relative value arb, event-driven, etc.) which tend to return, in aggregate, a small positive return in all regimes and then they lever it up. "60/40" essentially has two return streams - stocks and bonds. Some would argue it's more like 1.5 since stocks/bonds tend to follow each other more than not. So, yes, 60/40 isn't "well diversified" mathematically speaking.

[pathikrit]

Why this works, at a simplistic level, is multiple uncorrelated return streams

Actually, its just leveraged Harry Browne in there when you think about it. There are only 4 kinds of market:

• Prosperity / low rates: Growth stocks like QQQ beat SPY

• Recession / high rates: Just hold cash in short term CDs aka short-term T-bills (covered by KMLM)

• Deflationary growth / declining rates: Commodities (covered by KMLM)

• Inflation / rising rates: bonds (covered by KMLM)

And, for moments of uncertainty / geopolitical tensions / market panic add a shelter in Gold and something like $PSCC (small cap consumer staples):
30 year backtest: https://testfol.io/?d=eJzNkW1LwzAUhf9Ku ... I%2BIi%2Fe

[Mle12]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

[happenstance]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

Thanks for sharing. This is a great article explaining this asset class, and why funds in the same category can behave differently. I also appreciate this succinct explanation comparing managed futures to momentum, which I've struggled to understand:

Trend-following rhymes with momentum investing, a better-known philosophy. A couple of key features delineate the two. Momentum strategies tend to focus on stocks and show positive correlation to the stock market. Managed-futures funds can trade futures on stock, bond, commodity, and currency markets around the world. They are also willing to go long or short, resulting in low or negative correlation with the broad equity market, a key differentiator compared with long-only momentum ETFs.

[brightlightstonight]

I'm trying to understand sources of return (and losses) in these things before I consider taking the plunge. So a question about leverage in managed futures and short-term interest rates...

I'd gotten my head around how NTSX (a 90 equity/60 treasury ETF) works, and specifically that the leveraged part of it (the treasuries) has financing costs equivalent to short-term interest rates. So, the total performance is basically 90% S&P 500 + 60% treasuries - 50% cash (and at 50% rather than 60% because it holds 10% in cash). And so high short-term rates represent a meaningful performance headwind for that type of fund.

But after a lot more reading, I think that it doesn't work that way for Managed Futures. NTSX is all long, but Managed Futures are a mix of long and short. And the part of cost of carry which is short-term interest rates is a performance drag for long positions, but a performance win for short positions. Do I have that right?

If that's right, then (for example) KMLM is currently a bit over 200% gross leveraged if I add up all the exposures (as of August 16.). But if I add long and subtract short, I only get a net leverage of 31.62% (long-short), so with that exposure, it's got a ~32% cash drag from futures exposure (not 200%), and gets all that back from the cash and T-bills on hand. (Not that "performs somewhat worse than cash in a low-volatility market" is a selling point, but at least it's not a headwind.)

And obviously these numbers change over time as it moves between overall long and overall short depending on current trends - but is there any analysis of the average net exposure for any of these ETFs? Does it average out to 0, or is it typically net long?

Thanks in advance for any corrections - I'm coming from a place of ignorance, though CME courses (https://www.cmegroup.com/education/courses.html) have been really helpful.

[typical.investor]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

Interesting but ... as M* points out:

Trend-following taps a straightforward formula: Buy what’s up, sell what’s down. These systematic strategies eschew fundamental analysis, manager due diligence, and intuition, relying instead on investments’ recent direction to foretell their trajectory.

The best explanation for why trend-following works is that investors react to new information too slowly. Preconceived notions make us slow to incorporate positive news that should increase prices or bad news that should decrease them. Clients may be dubious of new information. Investment committees take time to review new theses. Systematic trend-following exploits the sloth, carving an advantage by capitalizing on human inefficiencies.

As more and more AI based systems try to address these inefficiencies, I wonder how the strategy will do going forward.

I like the fact that the returns are negatively correlated, but picking a good fund which will react in good time seems daunting. Perhaps a fund that replicates an index composed of the largest managed-futures strategies would be the best bet.

Still, the stock/bond correlation doesn't go positive very often. I am not convinced I truly have a need.

[whodidntante]

But after a lot more reading, I think that it doesn't work that way for Managed Futures. NTSX is all long, but Managed Futures are a mix of long and short. And the part of cost of carry which is short-term interest rates is a performance drag for long positions, but a performance win for short positions. Do I have that right?

That is correct for financing costs.

Physical commodity futures also have storage and insurance costs baked in, as well as financing. And folks like to speculate (ahem, price discovery) which influences the prices of long-dated contracts. This is referred to as backwardation/contango.

[Always passive]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

Interesting but ... as M* points out:

Trend-following taps a straightforward formula: Buy what’s up, sell what’s down. These systematic strategies eschew fundamental analysis, manager due diligence, and intuition, relying instead on investments’ recent direction to foretell their trajectory.

The best explanation for why trend-following works is that investors react to new information too slowly. Preconceived notions make us slow to incorporate positive news that should increase prices or bad news that should decrease them. Clients may be dubious of new information. Investment committees take time to review new theses. Systematic trend-following exploits the sloth, carving an advantage by capitalizing on human inefficiencies.

As more and more AI based systems try to address these inefficiencies, I wonder how the strategy will do going forward.

I like the fact that the returns are negatively correlated, but picking a good fund which will react in good time seems daunting. Perhaps a fund that replicates an index composed of the largest managed-futures strategies would be the best bet.

Still, the stock/bond correlation doesn't go positive very often. I am not convinced I truly have a need.

Not all managed futures strategies are based on trend following. There are hedge funds/managed futures that use multiple strategies. Look up CFM DISCUS operating successfully since 1991.

[comeinvest]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

Thanks for sharing. This is a great article explaining this asset class, and why funds in the same category can behave differently. I also appreciate this succinct explanation comparing managed futures to momentum, which I've struggled to understand:

Trend-following rhymes with momentum investing, a better-known philosophy. A couple of key features delineate the two. Momentum strategies tend to focus on stocks and show positive correlation to the stock market. Managed-futures funds can trade futures on stock, bond, commodity, and currency markets around the world. They are also willing to go long or short, resulting in low or negative correlation with the broad equity market, a key differentiator compared with long-only momentum ETFs.

I think momentum strategies usually refer too individual stocks.

[comeinvest]

Managed Futures: What to Know About This Strategy for ETFs | Morningstar

[Link formatted by admin LadyGeek]

Interesting but ... as M* points out:

Trend-following taps a straightforward formula: Buy what’s up, sell what’s down. These systematic strategies eschew fundamental analysis, manager due diligence, and intuition, relying instead on investments’ recent direction to foretell their trajectory.

The best explanation for why trend-following works is that investors react to new information too slowly. Preconceived notions make us slow to incorporate positive news that should increase prices or bad news that should decrease them. Clients may be dubious of new information. Investment committees take time to review new theses. Systematic trend-following exploits the sloth, carving an advantage by capitalizing on human inefficiencies.

As more and more AI based systems try to address these inefficiencies, I wonder how the strategy will do going forward.

I like the fact that the returns are negatively correlated, but picking a good fund which will react in good time seems daunting. Perhaps a fund that replicates an index composed of the largest managed-futures strategies would be the best bet.

Still, the stock/bond correlation doesn't go positive very often. I am not convinced I truly have a need.

Additionally, what is the impact of the surge in MF ETF and mutual fund assets following these strategies? What is the capacity of these strategies? How large are the combined assets of retail investors in these strategies in comparison to historical institutional money?
Google AI Overview and Bing Copilot have no clue.
It is clear that despite slight nuances between the individual implementations, all trend following money basically chases the same futures contracts at about the same time.
I can't find it googling right now, but didn't Goldman Sachs or some investment bank have huge funds following some commodities trade, that worked for a long time, until it didn't?

[kxl19]

Another relevant detail is that KMLM's expense ratio is 0.90%, and DBMF's is 0.85.

It is not a managed futures fund, but here is another fund that some would detest, BTAL. The inception of the etf is 2011 but I have historical data since 2001. It has lost money since its inception and its net expense is 1.43.

Is there a publicly accessible source for historical data that simulates the ETF? Where can I find it?

[brightlightstonight]

Another relevant detail is that KMLM's expense ratio is 0.90%, and DBMF's is 0.85.

It is not a managed futures fund, but here is another fund that some would detest, BTAL. The inception of the etf is 2011 but I have historical data since 2001. It has lost money since its inception and its net expense is 1.43.

Is there a publicly accessible source for historical data that simulates the ETF? Where can I find it?

testfol.io - use DBMFX and KMLMX.

[CletusCaddy]

Today is a great example of why this fund could be a core holding that diversifies the traditional three fund portfolio

[GaryA505]

Today is a great example of why this fund could be a core holding that diversifies the traditional three fund portfolio

What happened today?

[jdibber]

yep, and a little GLDM helps soften the ride