Do commodity futures funds have positive expected return?

[dyangu]

Do ETFs based on commodity future contracts (such as DBC) have positive expected return in the long term? These ETFs constantly buy and sell futures contracts and do not own physical commodities or shares in companies. How can they track the prices of the underlying commodities? Where does the expected return come from? Shouldn't the price of a futures contract already reflect market expectations, and therefore the expected return will just be the opportunity cost of paying upfront?

[Uninvested]

It is guaranteed. No other investments are needed. Just commodity ETFs.

[Quasimodo]

Here's a link to the Altruist website with a number of discussions about investing in commodities:

http://www.altruistfa.com/readingroomar ... ommodities

I'm not affiliated with Altruist Financial Advisors and only offer the link as a free research opportunity for those who might be interested in this type of investment.

John

[VictoriaF]

I am waiting for a bubble in comods. The problem is that "positive return" would require a lot of mopping.



Victoria

[staythecourse]

Do ETFs based on commodity future contracts (such as DBC) have positive expected return in the long term? These ETFs constantly buy and sell futures contracts and do not own physical commodities or shares in companies. How can they track the prices of the underlying commodities? Where does the expected return come from? Shouldn't the price of a futures contract already reflect market expectations, and therefore the expected return will just be the opportunity cost of paying upfront?

Wow a lot of dumping on a investment product, but no real intelligent answers.

You are correct in your logic. commodity etf should not have much real return. There are 3 sources of return on futures: Roll yield, spot price, and collateral return (tbills usually). Usually the 3 together give you a real return of tbills (likely even less if you take ER and transaction costs).

The idea of holding commodities is the diversification in improving risk/ return in a well diversified portfolio (which has been shown in the past). The other adv. should be a good hedge for rising inflation especially in peiods of HIGH inflation.

Mr. Swedroe in his book Right Financial Plan has an excellent 1-2 page summary of the adv. of holding commodities.

It is up to every intelligent investor to read and understand the different investment options out there and decide for oneself if they have a role or not in your portfolio.

Good luck.

[larryswedroe]

a) reasonable to assume commodities in long term move about in line with inflation. Could be higher or lower but IMO that is the reasonable assumption

b) let's also assume over time there is no contango or backwardation, or at least it is small.

c) you earn the collateral return which should have enough real return to offset expenses

d)any asset that tends to do well when risky assets do poorly should have low expected real return as they act as insurance

e) So net we should expect very low real return, if any, from individual commodities.

f) but you don't own individual commodities, or their futures. You own basket. So you get diversification return, shown to be several percent. As the funds rebalance. Similar to say owning 50/50 S&P/EAFE and you get higher returns than weighted average of the two, and sometimes you see even higher return than either, depending on the time period.So that helps

g) then the only RIGHT WAY to think about any asset is not in isolation, but how its addition impacts portfolio. The very low to negative correlation of CCF to stocks and bonds combined with their high volatility means you get a very large diversification return, perhaps as much as 4-5%, meaning CCF contributes returns to portfolio as if it really returned 4-5 more than it actually did--you can see that in the example in my book on alternative investments, and I have also presented that here as well.

h) the literature shows that construction matters, that you have in past been able to add value by managing rolls and also by choosing commodities to include based on roll costs--which is why I like the idea of the new Summerhaven funds and am considering investing.


So bottom line is that unless you think that correlations to stocks and bonds will all of a sudden rise dramatically, that correlations of individual commodities will rise dramatically and that volatility of CCF will fall dramatically, you should expect a higher portfolio benefit than just the individual return of the asset class, making CCF worth considering

I hope that helps

[raddle]

Wow a lot of dumping on a investment product, but no real intelligent answers.

Oh I get it! DUMPING! :lol:

[dyangu]

g) then the only RIGHT WAY to think about any asset is not in isolation, but how its addition impacts portfolio. The very low to negative correlation of CCF to stocks and bonds combined with their high volatility means you get a very large diversification return, perhaps as much as 4-5%, meaning CCF contributes returns to portfolio as if it really returned 4-5 more than it actually did--you can see that in the example in my book on alternative investments, and I have also presented that here as well.

Thanks for the reply, Larry. I understand the value of diversifying. However, I would not take 5% of my portfolio to a casino to diversify because the expected return is negative. I just want to understand where the expected return for commodity futures comes from.

[psteinx]

I'm no expert on commodities. That said, here's my understanding:

Expected return for commodities: 4 main parts.

1) Real return on the commodity itself. i.e. How should we expect the long term price trend of the types of commodities in these funds to go?

2) Return from the tendency of longer term commodities futures contracts to be consistently higher or lower than shorter term contracts (i.e. if the strategy is to buy 3 month contracts and sell them at the 1 month mark, rolling forward, what is the average price difference between 3 and 1 month contracts?)

3) Transaction costs of contracts, management fees and the like, and interest income if any from the collateral. i.e. The net gains and losses from the mechanics of implementing the strategy.

4) Diversification effects of implementing this strategy on a basket of commodities.

I don't have crystal clear estimates of what the right values for any of these should be. As I said, this isn't an area of expertise for me.

That said:

#1 is probably reasonably close to 0% real, long term. Maybe modestly positive (a percent or two) due to increased worldwide economic activity and resource depletion.

#2 is situational. Depends to a large extent on the supply and demand for either side of the contracts, from producers, consumers (industrial), speculators/investors and the like. These values have likely changed over time. What held in the 1970s likely did not in the 1990s, and the increasing popularity of commodity funds in the last 3-4 years or so has likely caused meaningful shifts in supply/demand relationships.

#3 - To some extent, you should be able to look at the vehicles you're interested in and figure this out. Hopefully your target vehicle has some history and you can compare how it tracked versus a close index over that time. Of course, these things vary over time, so figures for 2008 may not translate well to 2011. In particular, short term interest rates swing quite a bit.

#4 Complicated issue, IMO. If commodity prices (or returns, maybe) are mean reverting and not closely correlated, then the long term performance of a rebalanced basket should exceed that of the individual components. If prices/returns are truly random, then it is not so clear. Would probably need to do a big spreadsheet analysis and post it here to be clearer about what I mean.

Disclosure: Though I'm not a big fan of commodity investment strategies, I do in fact have a significant position in SKMRX, which is in this ballpark. The position is a result of a CEF (closed end fund) play I made and I will likely exit the position sometime between now and the end of this year.

[psteinx]

One quick follow up since I mentioned indexes in my previous post.

Be very careful about using a quoted historical return on a commodities index to evaluate the strategy.

It's my understanding that many of the popularity commodity indexes were constructed relatively recently but then backtested to create longer historical data.

If those who choose the rules to construct an index know that those rules are likely to be used to backfill historical results from prior to the development of those rules, then there's a strong incentive to design the rules in a way so as to maximize historical returns. There are lots of little judgement calls in creating a ruleset - especially for something a bit complicated to design such as a commodities index based on futures contracts, and so plenty of chances to tilt those choices to make the historical track record look good.

If one is only looking at returns data AFTER the index was developed, this is less of a problem (or perhaps no problem at all). But I'm not sure which, if any, commodities indexes were developed and tracked far enough back (1970s or before, preferably), to create a nice long track record in real time.

[larryswedroe]

I have showed many times the benefits of adding CCF have produced when added to a portfolio--basically add a few percent and you keep returns unchanged but reduce risk. And because of high volatility and low correlation you don't need much. Also helps to cut tail risks and can also add some bond duration due to their being negatively correlated to bonds. And my examples did not even include this benefit. But that shows that it is foolish to consider assets in isolation (an issue that is not even debated in academia)

I hope that is helpful

[staythecourse]

I have showed many times the benefits of adding CCF have produced when added to a portfolio--basically add a few percent and you keep returns unchanged but reduce risk. And because of high volatility and low correlation you don't need much. Also helps to cut tail risks and can also add some bond duration due to their being negatively correlated to bonds. And my examples did not even include this benefit. But that shows that it is foolish to consider assets in isolation (an issue that is not even debated in academia)

I hope that is helpful

Not considering an asset allocation in isolation is discussed ALL THE TIME, but most folks still spend all their effort trying to figure out what changes in interest rates will do to their bond portion, how do I get some more yield from my cash, is large vs. small cap the best, etc. As you scan this site alone most talk about assets in isolation (and the factors affecting them), so despite being well known not to view assets in isolation people don't seem to focus on what they preach.

Good luck.

[larryswedroe]

I agree completely

[Wagnerjb]

But that shows that it is foolish to consider assets in isolation (an issue that is not even debated in academia)

Larry is correct about an individual looking at assets and how they interact with other parts of your portfolio. Markets do that too. For example, an asset with negative correlation to stocks will be priced (by market participants) with very low returns. That is because that asset has beneficial interaction with stocks in your portfolio, so people are willing to accept lower returns for that particular asset. Gold is an example of an asset with slight negative correlation to stocks, and a commensurately low return.

You should look at how assets interact with other parts of your portfolio, because the market prices assets that way too.

Best wishes.

[larryswedroe]

Exactly, Which is why assets that act as portfolio insurance should have very low expected returns. And in times of high risk the expected returns should go even lower (prices higher) because investors are willing to pay even larger risk premiums for portfolio insurance. That IMO likely explains at least a significant part of the run up in the price of gold, as well as why Treasuries have rallied despite Gross's forecast and the forecast of most others.

See here for more comments on this issue

http://moneywatch.bnet.com/investing/bl ... blog-river

[FinsterF]

Categorical statements of the form “X has a (low or high or whatever) return” are always fraught with peril unless made with reference to a particular time frame. And unless it’s a past time frame, they involve forecasting. There’s really no such thing as “is” returning anything.

The example of gold highlights this. Although it was also true in 2011, it remains true in 2024 that gold has exceeded the returns on stocks century to date. Whatever theory divides assets into categories and then assigns one to be an intrinsically low or high return asset on that basis lacks validity.