Commodities: Positive Returns = Inconclusive

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Jerry_lee
Posts: 915
Joined: Tue Mar 06, 2012 1:55 pm

Commodities: Positive Returns = Inconclusive

Post by Jerry_lee » Wed May 02, 2012 1:56 pm

Thought some would find this interesting: we have data on the DJ-AIG Commodity index dating back to 1991, now over two decades. While the CCF return has been positive (+5.3% through April of 2012), the problem is, there are two components to commodity index returns: the commodity returns themselves, and the collateral return, 3mo t-bills in the case of DJ-AIG.

So to measure whether there is an actual return in excess of the collateral (which we know has a positive return), we can check to see if there is a statistically significant positive return for the DJ-AIG Index above 3mo t-bills. What we find is, while the DJ-AIG has bested 3mo t-bills by 0.24% per month, the t-stat is only 0.88, meaning that we cannot say with any confidence that the return of commodities is any higher than the collateral itself.

Conversely, by stepping out just a few more months in the fixed income market, we see that some additional term risk has provided reliably higher returns: +0.02% per month for 6mo t-bills and +0.05% for 1yr t-notes, with t-stats nearing 7 and 5 (highly significant).

But short term bonds have very low volatility, so it is easy to determine return premiums. Is it only the high volatility of CCFs (SD of 15) that create the question of returns? No. The Russell 3000 Index had a very similar SD (15.3), outperformed 3mo t-bills by 0.56% per month, with a t-stat of 2.0, indicating a conclusive premium over the risk free rate.

Be very wary of "investments" with questionable return properties. Probably better sticking with plain 'ole stocks (small and value in particular) and bonds (short term high quality), really hard to go wrong.
The most disciplined investor in the world.

hazlitt777
Posts: 1033
Joined: Fri Aug 12, 2011 4:10 am
Location: Wisconsin

Re: Commodities: Positive Returns = Inconclusive

Post by hazlitt777 » Wed May 02, 2012 2:08 pm

What do you mean by the "collateral return." What are you referring to?

User avatar
oneleaf
Posts: 2344
Joined: Mon Feb 19, 2007 5:48 pm

Re: Commodities: Positive Returns = Inconclusive

Post by oneleaf » Wed May 02, 2012 4:03 pm

Thanks for posting this, it is very interesting.

i stopped including a substantial CCF portion. I think, by the nature of these instruments, there would be zero excess return beyond collateral return unless there is a roll return. The only other return that can be a solid source of return is that accomplished by extending the duration of your bond portfolio, which can be done if CCF's are used to hedge the term risk. Considering the high costs, I decided against CCF's.

User avatar
magician
Posts: 1568
Joined: Mon May 02, 2011 1:08 am
Location: Yorba Linda, CA
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by magician » Thu May 03, 2012 1:17 am

Jerry_lee wrote:. . . there are two components to commodity index returns: the commodity returns themselves, and the collateral return, 3mo t-bills in the case of DJ-AIG.
There is a third component: roll yield. As futures (or forward) contracts expire you have to roll them into new futures (or forwards). If markets are in contango, the new contracts will have a higher price than the old contracts, so the roll yield will be negative. If markets are in backwardation, the new contracts will have lower prices than the old contracts, so the roll yield will be positive.
Simplify the complicated side; don't complify the simplicated side.

User avatar
magician
Posts: 1568
Joined: Mon May 02, 2011 1:08 am
Location: Yorba Linda, CA
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by magician » Thu May 03, 2012 1:21 am

hazlitt777 wrote:What do you mean by the "collateral return." What are you referring to?
When you enter into a futures contract, you are required to post collateral (most people post Treasury securities) in case the market price moves adversely. The unused collateral earns interest; that interest is called the collateral yield (or collateral return).
Simplify the complicated side; don't complify the simplicated side.

User avatar
oneleaf
Posts: 2344
Joined: Mon Feb 19, 2007 5:48 pm

Re: Commodities: Positive Returns = Inconclusive

Post by oneleaf » Thu May 03, 2012 1:49 am

magician wrote:
Jerry_lee wrote:. . . there are two components to commodity index returns: the commodity returns themselves, and the collateral return, 3mo t-bills in the case of DJ-AIG.
There is a third component: roll yield. As futures (or forward) contracts expire you have to roll them into new futures (or forwards). If markets are in contango, the new contracts will have a higher price than the old contracts, so the roll yield will be negative. If markets are in backwardation, the new contracts will have lower prices than the old contracts, so the roll yield will be positive.
I think the roll yield is what is considered the "the commodity returns themselves", since the roll yield IS the return from the commodities for the futures contract holder. If "the commodity returns themselves" refers to expected spot return, that would be false. The CCF holder is not exposed to the spot return because any expected movement is priced into the contract. So only if Keyes and Hicks' theory of normal backwardation is true would there be a risk premium accrueing for the buyer, on average.

hazlitt777
Posts: 1033
Joined: Fri Aug 12, 2011 4:10 am
Location: Wisconsin

Re: Commodities: Positive Returns = Inconclusive

Post by hazlitt777 » Thu May 03, 2012 7:50 am

magician wrote:
hazlitt777 wrote:What do you mean by the "collateral return." What are you referring to?
When you enter into a futures contract, you are required to post collateral (most people post Treasury securities) in case the market price moves adversely. The unused collateral earns interest; that interest is called the collateral yield (or collateral return).
Thanks. Learn something everyday.

User avatar
magician
Posts: 1568
Joined: Mon May 02, 2011 1:08 am
Location: Yorba Linda, CA
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by magician » Thu May 03, 2012 10:47 am

oneleaf wrote:
magician wrote:
Jerry_lee wrote:. . . there are two components to commodity index returns: the commodity returns themselves, and the collateral return, 3mo t-bills in the case of DJ-AIG.
There is a third component: roll yield. As futures (or forward) contracts expire you have to roll them into new futures (or forwards). If markets are in contango, the new contracts will have a higher price than the old contracts, so the roll yield will be negative. If markets are in backwardation, the new contracts will have lower prices than the old contracts, so the roll yield will be positive.
I think the roll yield is what is considered the "the commodity returns themselves", since the roll yield IS the return from the commodities for the futures contract holder. If "the commodity returns themselves" refers to expected spot return, that would be false. The CCF holder is not exposed to the spot return because any expected movement is priced into the contract. So only if Keyes and Hicks' theory of normal backwardation is true would there be a risk premium accrueing for the buyer, on average.
Not exactly, but close.

There is a spot yield (the difference between the spot price at the expiration of the contract and the spot price at the inception of the contract) and a roll yield (the difference between two differences: the change in the futures price less the change in the spot price); it appears that what you're calling the roll yield is the sum of those two.

I don't know what, exactly, the OP meant by "the commodity returns themselves": he may have meant the spot yield, or he may have meant the sum of the spot yield and the roll yield; as you say, the latter seems more likely.
Simplify the complicated side; don't complify the simplicated side.

staythecourse
Posts: 6118
Joined: Mon Jan 03, 2011 9:40 am

Re: Commodities: Positive Returns = Inconclusive

Post by staythecourse » Thu May 03, 2012 10:53 am

Once again another person makes a fundamental problem in analyzing assets for a portfolio. NO ONE is proposing or has proposed of 100% commodities. Have they??

The more I see of investors the more I am starting to realize some get the fact it is the portfolio as a whole that matters and not its individual components and some don't. So the question of SCV, iSV, treasuries, EM bonds, etc.. should not be looked at in isolation, but as part of a portfolio. Markowitz wrote about in in the 50's and won a Nobel Prize for these observations and folks still don't seem to get it.

If you like commodities or not is not the point. The analysis that should be done is does its addition help or hurt a portfolio as a whole.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

User avatar
oneleaf
Posts: 2344
Joined: Mon Feb 19, 2007 5:48 pm

Re: Commodities: Positive Returns = Inconclusive

Post by oneleaf » Thu May 03, 2012 11:50 am

magician wrote: There is a spot yield (the difference between the spot price at the expiration of the contract and the spot price at the inception of the contract) and a roll yield (the difference between two differences: the change in the futures price less the change in the spot price); it appears that what you're calling the roll yield is the sum of those two.

I don't know what, exactly, the OP meant by "the commodity returns themselves": he may have meant the spot yield, or he may have meant the sum of the spot yield and the roll yield; as you say, the latter seems more likely.
I always find the term "spot yield" confusing since the only exposure to spot price movements is that which was unexpected. The difference between roll yield and spot yield depends on knowing exactly what the expected returns were at the time the contract was set, which is unknowable both ex ante and ex post, unless a survey was used to ask every participant (this is true of all asset classes, where expected return is never knowable and is only one part of the realized return).

User avatar
oneleaf
Posts: 2344
Joined: Mon Feb 19, 2007 5:48 pm

Re: Commodities: Positive Returns = Inconclusive

Post by oneleaf » Thu May 03, 2012 12:41 pm

staythecourse wrote:Once again another person makes a fundamental problem in analyzing assets for a portfolio. NO ONE is proposing or has proposed of 100% commodities. Have they??

The more I see of investors the more I am starting to realize some get the fact it is the portfolio as a whole that matters and not its individual components and some don't. So the question of SCV, iSV, treasuries, EM bonds, etc.. should not be looked at in isolation, but as part of a portfolio. Markowitz wrote about in in the 50's and won a Nobel Prize for these observations and folks still don't seem to get it.

If you like commodities or not is not the point. The analysis that should be done is does its addition help or hurt a portfolio as a whole.

Good luck.

To be fair, the OP is questioning the positive return characteristics of commodities, which has been touted as positive (or even equity-like, by some). If studies prove otherwise, it is relevant to portfolio design. For instance: low correlation AND equity-like returns make them attractive to many. Low correlation but no return only makes it attractive for those that want or need the "insurance" property of it.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Thu May 03, 2012 12:49 pm

all of these miss the major point and is classic mistake, thinking of assets in isolation and not how their addition impacts risk and return of the whole portfolio.
The diversification return of an asset with negative to low correlation and very high volatility is very high.
In my books I showed examples that the contribution of CCF to a portfolio was several percentages above return of the asset itself, as much as 5%.
Larry

Bongleur
Posts: 2066
Joined: Fri Dec 03, 2010 10:36 am

Re: Commodities: Positive Returns = Inconclusive

Post by Bongleur » Thu May 03, 2012 9:13 pm

>In my books I showed examples that the contribution of CCF to a portfolio was several percentages above return of the asset itself, as much as 5%.
>

I'd like to understand if that result is dependent upon the amount of time spent in contango vs normal backwardation. It would be reassuring to know that it worked in an extended period of one or the other.

EDIT: does flipping between contango and backwardation account for much of the volatility, or is it "volatile enough" without?
Seeking Iso-Elasticity. | Tax Loss Harvesting is an Asset Class. | A well-planned presentation creates a sense of urgency. If the prospect fails to act now, he will risk a loss of some sort.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Fri May 04, 2012 7:42 am

Bongleur

This is from my alternative investment book and is good example of the benefits

Consider the following results, covering the period from 1991 through 2007.
Portfolio Allocation Annualized Return Standard Deviation
100% DJ-AIG 7.91% 16.63%
100% GSCI 6.80% 25.62%
100% S&P 500 11.41% 17.00%
95% S&P 500/5% GSCI 11.42% 15.94%
95% S&P 500/5% DJ-AIG 11.39% 15.98%
There are two important observations. First, while both the DJ-AIGCI and the S&P GSCI underperformed the S&P 500 Index by large amounts, the addition of a 5 percent allocation to either resulted in a more efficient portfolio.

You can obviously see that the IMPACT of adding say the GSCI was not as if it earned 6.8% but as if it earned about 4.5% more than that!!! Great example of why one should never consider asset in isolation.

Personally I prefer the DJ AIG because it is more diversified and also minimizes contango issues because not so heavily in energy. In the easily stored commodities it's all about the arbitrage, so not a real issue. Also I would add there is strong live evidence that one can outperform the benchmarks which are easily exploited because of front running against indexers who must trade on specific dates. This can be as much as a couple of percent per year (which is about what the Russell problem was in size) and you can also improve results by not trading always the front month, but the cheapest to roll

I hope this is helpful

Larry

555
Posts: 4955
Joined: Thu Dec 24, 2009 7:21 am

Re: Commodities: Positive Returns = Inconclusive

Post by 555 » Fri May 04, 2012 8:24 am

Larry, the BlackRock LifePath Index funds, e.g. link
http://www.oregon.gov/PERS/OSGP/docs/Li ... S.pdf?ga=t (12 page PDF)
have recently added a 3%-4% allocation to a Commodity fund called DJ-UBS Commodity Daily Fd E.
Do you have any thoughts on if this is a good idea?
Or if BlackRock will use it advantageously (as opposed to window dressing)?

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Fri May 04, 2012 9:07 am

Unfortunately for CCFs, everything Larry says about commodities applies even better to owning gold (portfolio diversification benefit). And yet we are told that investing in gold is only for people who wear tin hats.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

staythecourse
Posts: 6118
Joined: Mon Jan 03, 2011 9:40 am

Re: Commodities: Positive Returns = Inconclusive

Post by staythecourse » Fri May 04, 2012 9:50 am

Lbill wrote:Unfortunately for CCFs, everything Larry says about commodities applies even better to owning gold (portfolio diversification benefit). And yet we are told that investing in gold is only for people who wear tin hats.
100% agree. That is the reason I always advocate get as much information as possible then each person needs to make their own judgements on what and what not to include in your portfolio.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Fri May 04, 2012 10:20 am

Lbill
Not true at all. While I have no problem with people owning an allocation to gold, gold DOES NOT HEDGE inflation it is not a factor in inputs to inflation.
Thus IMO broader based CCF is a better alternative, but to each his own.
Best wishes
Larry

555
Posts: 4955
Joined: Thu Dec 24, 2009 7:21 am

Re: Commodities: Positive Returns = Inconclusive

Post by 555 » Fri May 04, 2012 10:27 am

What is DJ-UBS Commodity Daily Fd E? (See my post above.)

User avatar
camontgo
Posts: 324
Joined: Tue Oct 27, 2009 1:46 pm
Location: Chicago, IL
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by camontgo » Fri May 04, 2012 10:51 am

Jerry_lee wrote:So to measure whether there is an actual return in excess of the collateral (which we know has a positive return), we can check to see if there is a statistically significant positive return for the DJ-AIG Index above 3mo t-bills. What we find is, while the DJ-AIG has bested 3mo t-bills by 0.24% per month, the t-stat is only 0.88, meaning that we cannot say with any confidence that the return of commodities is any higher than the collateral itself.
I agree that the data is inconclusive (fwiw, i'm not a big fan of CCFs myself), but 20 years isn't much data in the volatile world of finance.

Even the equity risk premium would very often have a t-stat of less than 2 using 20 years of annual return data. Fama-French data shows the equity risk premium from 1926-2011 has a Sharpe ratio of 0.38, which suggests that on average we need about 27 years to get a t-stat above 2.

I'd say the stats are "non-surprisingly inconclusive given the limited data".
"Essentially, all models are wrong, but some are useful." - George E. P Box

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Fri May 04, 2012 10:54 am

larryswedroe wrote:Lbill
Not true at all. While I have no problem with people owning an allocation to gold, gold DOES NOT HEDGE inflation it is not a factor in inputs to inflation.
Thus IMO broader based CCF is a better alternative, but to each his own.
Best wishes
Larry
One tactic of skilled debaters is to shift the topic: so now we're on the subject of inflation hedging instead of the original topic. Couple points:
> If you want an inflation hedge, buy TIPs. Neither gold nor commodities are as effective.
> Not everyone regards "inflation" as increases in the consumer price index. Many people regard inflation as increases in the money supply, and gold is an effective hedge for that.
>There's a big difference between the spot price of commodities and the returns of commodity futures. There's a shocking amount of tracking error between rising commodity prices and the prices of commodity futures. Futures don't provide any better hedge for CPI inflation than gold over time; the only "proof" offered for this argument is based on cherry-picking the time period starting in 1981 right at the peak of the gold price run-up in the 1970s when the Fed dramatically raised interest rates to double digits kill high inflation. That was a one-off event that will never, ever be repeated in our lifetimes - if you think it will, then I have a bridge I'd be interested in selling you.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Fri May 04, 2012 11:32 am

Even though gold (and commodities) are not good short-term CPI inflation hedges, the price of gold is correlated 0.48 with CPI-U over the time period 1972-2011 (Simba), while commodity futures (DJ-AIG) is correlated 0.22, base on annual returns. So there's no basis for claiming that gold is not as good an inflation hedge as CCFs, IMO. This long term correlation of annual returns with CPI is confirmed in a paper by McCown & Zimmerman

In addition, the authors found there was statistically significant evidence at the .05% level of co-integration between the price of gold and consumer prices; i.e., that there is evidence of a common stochastic trend.
Abstract:
Gold shows the characteristics of a zero-beta asset. It has approximately the same mean return as a Treasury Bill and bears no market risk. Silver also bears no market risk but has returns inferior to Treasury Bills. Both gold and silver show evidence of inflation-hedging ability, with the case being much stronger for gold. The prices of both metals are cointegrated with consumer prices, showing additional evidence of hedging ability.
There is simply no convincing evidence that gold is not as good a hedge for CPI inflation as commodity futures; the only "proof" offered up is the fact that gold didn't keep pace with CPI during the cherry-picked time period beginning in 1981, which was a one-off time series anyway. This said, neither one is as good an inflation "hedge" as either T-Bills or TIPs, so this is a straw man argument in the first place.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
Rick Ferri
Posts: 8594
Joined: Mon Feb 26, 2007 11:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by Rick Ferri » Fri May 04, 2012 11:55 am

Unlike some, I don't rely on the "portfolio effect" to create returns. It's nice if it happens, but not a good strategy, IMO.

My philosophy has always been to invest in assets classes that:

1) Are fundamentally different than other asset classes.
2) Have a strand-alone expected real return.
3) Pay income in the form of dividends, interest or rents.
4) Can be invested in using very low cost funds.

This cuts out commodities. They do not produce an inflation-adjusted return in isolation (per Larry's explanation) and they do not provide positive cash flow in a portfolio. Commodity funds are also much more expensive than equity index funds. Larry is correct that at times commodities reduce the volatility in a portfolio, but taking away from stocks to invest in them also reduce long-term expected portfolio returns. You can't eat lower volatility.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Fri May 04, 2012 4:11 pm

Lbill
First, as I said I have no objection to using gold, we just disagree on which is preferable
Second, the data on correlation of commodities to inflation is that
a) the longer your term the higher the correlation (which is very logical, given the high volatility and that commodities are source of inflation)
in my book on alternatives I show that the correlation of CCF to inflation was 0 at monthly level, .14 at quarterly, .29 at annual (stat significant) and .45 at five years.
b) However, commodities tend to do well during RISING inflation and not well went it is falling.

Gold fell about 70% in value while inflation was running at 4% a year for about 25 years, so IMO that rules out gold as inflation hedge. We don't see that with commodities.

And inflation is not an increase in the money supply, but an increase in prices. You can have MS rise but velocity fall so you don't see the price impact.

Gold does have some geopolitical risks better than commodities and commodities hedge supply disruptions better than gold.

Also for a diversified portfolio of commodities that is not dominated by energy you don't have the contango problem as a big issue, especially if trade around it. And with CCF you can also add value with very limited risks by altering the collateral. Simply going out to one year or two can add 1-2% a year to returns. And can use TIPS which allows you to earn the term premium without the inflation risk you would have with nominals.

so take your pick. I prefer CCF

As to Rick's comments, obviously we disagree and that is because IMO he makes the mistake of thinking in isolation, failing to consider the diversification benefits, ignoring them IMO is wrong. But again, take your pick


Best wishes
Larry

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Fri May 04, 2012 6:24 pm

Gold fell about 70% in value while inflation was running at 4% a year for about 25 years, so IMO that rules out gold as inflation hedge. We don't see that with commodities.
From 1981-2005 (25 years) gold did have a compound annual real return of -4%, so it did lag CPI on average by 4% per year. Interest rates have fallen since 1981, from about 15% on the 10-year to about 2% currently. Gold will not do well during a period of falling real interest rates, while stocks, bonds, and commodities will. An important question for investors is where real interest rates are headed from here. Are we likely to see another quarter century of falling rates? If so, then stocks, bonds, and commodities will do well. If not, then maybe not so much. A problem with commodities is that they tend to go up during secular cycles when stocks and bonds are going up too, as in the period from 1981-2000. Gold seems to do a better job of diversifying stocks, so it's no surprise that the gold price moves contrary to stock returns, in both directions. If you buy gold as a diversifier, then don't ask for sympathy if stocks are doing well and your gold is dead money. But you'll be the one laughing on the way to the bank when stocks are tanking and commodities are probably flopping right along with them.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
Rick Ferri
Posts: 8594
Joined: Mon Feb 26, 2007 11:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by Rick Ferri » Fri May 04, 2012 9:02 pm

Larry Swedrow wrote:As to Rick's comments, obviously we disagree and that is because IMO he makes the mistake of thinking in isolation, failing to consider the diversification benefits, ignoring them IMO is wrong. But again, take your pick
I was very clear in addressing the issue of portfolio effect (diversification benefit) by stated that I wouldn't rely on this for excess return. I would rather rely on cash flows from dividends, interest and rent, and the real growth from stocks and REITs than the hope that higher cost no-growth /no income commodity funds deliver a portfolio effect in the future.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Fri May 04, 2012 10:04 pm

If one doesn't rely on the diversification benefit then one is ignoring it. Same thing.

Here is another example. Say you think stocks will return 7% and small stocks will have a 2% premium, but you don't want to count on that. So instead you "conservatively" estimate a 1% premium and forecast 8% for the asset class. When you do this you now have to hold a higher equity allocation than you would have if you did assume the 2% premium which you believed was the more accurate figure, and that causes you to take more risk than you would overall.

These kinds of decisions, like being "conservative" in estimating returns and thinking of asset classes in isolation (Markowitz won Nobel Prize for showing that was wrong) leads to suboptimal decisions.

Best wishes
Larry

User avatar
brick-house
Posts: 475
Joined: Thu May 07, 2009 1:16 pm
Location: Philadelphia, PA

Re: Commodities: Positive Returns = Inconclusive

Post by brick-house » Sat May 05, 2012 6:58 am

rick ferri wrote:
You can't eat lower volatility.
The School of Hard Knocks recently commissioned a series of studies. The studies showed that a diet high in expected returns can lead to severe heart problems in periods of extreme volatility or secular bear markets. Further findings showed that diets high in expected returns promote harmful behaviors such as panic selling and under-saving.

An interesting subset of the data showed a small group of investors who ate a diet heavy in low volatility and low expectations. The researchers found that low volatility and lowered expectations did wonders for mood and promoted positive behaviors like over-saving, disciplined re-balancing, low costs, and tax efficiency.

The studies recommended focusing on your diet as a whole, not on each food group in isolation. An example was given of a stock-obese investor whose health radically improved when gold and long term treasuries were added to the diet. :beer
You don't need no gypsy to tell you why- Greg Allman

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 8:23 am

Unfortunately the statement you cannot eat lower volatility is incorrect. That is only true in isolation. Not in portfolio context. Lower volatility takes annual average returns and improves compound returns of a portfolio. You get that from assets with negative to very low correlation and high volatility. Best wishes Larry

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 8:27 am

Lbill
IMO you have it exactly backwards
Gold will likely do poorly in periods of RISING real interest rates and should do well in periods of falling real rates. In fact it is the very low real rates that Greenspan imposed and now Bernanke that led to Gold's rise.
The cost of alternatives to gold (lost interest) becomes less. In rising real rates the cost of hold gold increases, lowering demand. Also rising real rates is inflation fighter and vice versa. Same true for commodities in general. In fact the research shows that commodities do well in RISING inflation and poorly in FALLING inflation

Best wishes
Larry

aac74
Posts: 64
Joined: Mon Apr 30, 2012 2:27 pm

Re: Commodities: Positive Returns = Inconclusive

Post by aac74 » Sat May 05, 2012 8:37 am

You have to consider government debt. In the US it's 100% of GDP !

How do you raise rates to the point that it kills the commodity bull market (like early 80s) without triggering a US debt default / end the the freaking world ?

What will the US look like after a debt default and austerity that follows ???

http://2.bp.blogspot.com/-sahWsw2yQ6M/T ... ax-Max.jpg

User avatar
NoRoboGuy
Posts: 732
Joined: Fri Apr 01, 2011 11:07 pm
Location: Just west of Austin TX

Re: Commodities: Positive Returns = Inconclusive

Post by NoRoboGuy » Sat May 05, 2012 9:04 am

If the diversification benefit alone makes an investment worthy of addition to a portfolio, then I can think of many other "alternative" investments that would provide a great correlation benefit: coins, lottery tickets, African beads, baseball cards...etc.

I think it is quite reasonable to look at investments in isolation for purposes of considering whether they generate a real return, and then consider the correlation benefit. I can even understand holding a small portion of gold or commodities, small because the correlation benefit might outweigh the portfolio-level reduction in real return.

I think of the entire portfolio but I prefer all of its parts generating a real contribution over long periods, not just from rebalancing and volatility reduction.
There is no free lunch.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 9:17 am

LouisC
Of course the diversification ALONE should not be the only consideration. You'll note I stated that one should consider how the addition of the asset impacts the ENTIRE portfolio

Also classic economic theory tells you that assets that do poorly in bad times (when other portfolio assets are doing poorly) should have very low returns because they act like insurance. So one should in isolation expect very low real returns. But we still buy insurance even though we estimate we will have negative return on the purchase.

I hope that is helpful

Best wishes
Larry

Bongleur
Posts: 2066
Joined: Fri Dec 03, 2010 10:36 am

Re: Commodities: Positive Returns = Inconclusive

Post by Bongleur » Sat May 05, 2012 9:54 am

larryswedroe wrote:Bongleur
This is from my alternative investment book and is good example of the benefits
Consider the following results, covering the period from 1991 through 2007.
snip
There are two important observations. First, while both the DJ-AIGCI and the S&P GSCI underperformed the S&P 500 Index by large amounts, the addition of a 5 percent allocation to either resulted in a more efficient portfolio.

You can obviously see that the IMPACT of adding say the GSCI was not as if it earned 6.8% but as if it earned about 4.5% more than that!!!
Larry
What happened in 2008-9? Did the inclusion of CCF reduce the losses or magnify them? Of course, if one had been doing it for many years the size of the portfolio was larger; so even if % losses were larger, the remaining portfolio size might still be larger than if 100% equities.

***
Its really hard to talk about gold. You have to justify your starting point because the market has been altered so many times -- the change from silver to gold standard in the US in the 19th century (which moved silver to Asia, affecting that economy...) ; the early 20th century peg to gold, Roosevelt stealing the gold; Nixon selling it back & the creation of pure fiat USD...

LBill and Larry draw completely opposite conclusions about the direction of gold during inflation & deflation...
Seeking Iso-Elasticity. | Tax Loss Harvesting is an Asset Class. | A well-planned presentation creates a sense of urgency. If the prospect fails to act now, he will risk a loss of some sort.

Bongleur
Posts: 2066
Joined: Fri Dec 03, 2010 10:36 am

Re: Commodities: Positive Returns = Inconclusive

Post by Bongleur » Sat May 05, 2012 9:56 am

Oh, & noticed in Larry's Alt Invest book that the PIMCO CCF fund has a much cheaper share class thru advisors -- about half a percent cheaper; so that can largely make up for an AUM advisor fee.
Seeking Iso-Elasticity. | Tax Loss Harvesting is an Asset Class. | A well-planned presentation creates a sense of urgency. If the prospect fails to act now, he will risk a loss of some sort.

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Sat May 05, 2012 10:24 am

There seem to be three time periods in which gold and commodities acted differently relative to stocks and bonds: 1972-1980 (inflationary), 1981-2000 (deflationary), and 2001-2011 (mixed). The first of these was one of rising inflation and interest rates. Stocks and bonds fared poorly, while gold and commodities had positive returns. The second period was one of declining inflation and interest rates. Stocks and bonds both did well, commodities also had positive returns, while gold declined. In the third period, stocks have been flat, bonds have had positive returns, and both gold and commodities have had positive returns.

When we examine the correlation between commodities and gold with the CPI-U during these periods, we find some interesting results. Here are the correlations between the real (inflation adjusted) returns and the CPI during each of these periods:

Period, Commodities, Gold
1972-1980, -0.13, 0.46
1981-2000, -0.08, -0.49
2001-2011, 0.52, 0.25

There are some unexpected data here. First, we see that the real returns of commodities were actually negatively correlated to inflation in the inflationary 1970s, while the real returns of gold were positively correlated. So gold actually was a more effective inflation hedge than commodities during this period.

Second, during the second period of declining or steady inflation, the real returns of commodities had essentially a zero correlation to inflation, while the correlation of the real returns of gold with inflation were actually negative. It's not so much that commodities were an effective inflation hedge during the 1981-2000 time period as it was that gold was actually a negative hedge. The claim that commodities are an effective inflation hedge is not supported for either the 1970s or the 1981-2000 period.

During the 2001-2011 period, the real returns of both gold and commodities were positively correlated to the CPI. Strangely enough, this is the only time period in which commodities were an effective inflation hedge, but so was gold. I guess one takeaway is that it's probably not good to own gold during a deflationary period that follows a period of high and accelerating inflation. However, we have to get through the period of high and accelerating inflation first, and for that you are probably better off owning gold than commodities. Be careful out there...
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
NoRoboGuy
Posts: 732
Joined: Fri Apr 01, 2011 11:07 pm
Location: Just west of Austin TX

Re: Commodities: Positive Returns = Inconclusive

Post by NoRoboGuy » Sat May 05, 2012 11:08 am

larryswedroe wrote:LouisC
Of course the diversification ALONE should not be the only consideration. You'll note I stated that one should consider how the addition of the asset impacts the ENTIRE portfolio

Also classic economic theory tells you that assets that do poorly in bad times (when other portfolio assets are doing poorly) should have very low returns because they act like insurance. So one should in isolation expect very low real returns. But we still buy insurance even though we estimate we will have negative return on the purchase.

I hope that is helpful

Best wishes
Larry
Thanks. Maybe I just do not understand the math supporting the re-allocation from bonds to commodities to "insure." If both serve as insurance, what is the incremental benefit if one (bonds) has a real return and the other (commodities) does not? Here is a Vanguard paper that relates to this discussion:
Because bonds have relatively low volatility in
addition to low average correlations to stocks,
investors have traditionally used bonds to diversify
their stock allocations. However, investment
products such as exchange-traded funds (ETFs)
have arisen in recent years, providing simplified,
low-cost access to a greater number of risk-premium
asset classes and sub-asset classes beyond U.S.
bonds. As a result, it’s no surprise that attention has
been drawn to the potentially diversifying properties
of investments such as commodities, real estate,
emerging-market bonds, and micro-cap stocks, to
name a few. Academic research and historical
experience suggest that many of these higher-risk,
yet potentially diversifying, assets may provide
returns higher than those available in a typical
bond portfolio, even as they have been relatively
uncorrelated to U.S. stocks and bonds.
There is no free lunch.

User avatar
LH
Posts: 5490
Joined: Wed Mar 14, 2007 2:54 am

Re: Commodities: Positive Returns = Inconclusive

Post by LH » Sat May 05, 2012 12:32 pm

staythecourse wrote:Once again another person makes a fundamental problem in analyzing assets for a portfolio. NO ONE is proposing or has proposed of 100% commodities. Have they??

The more I see of investors the more I am starting to realize some get the fact it is the portfolio as a whole that matters and not its individual components and some don't. So the question of SCV, iSV, treasuries, EM bonds, etc.. should not be looked at in isolation, but as part of a portfolio. Markowitz wrote about in in the 50's and won a Nobel Prize for these observations and folks still don't seem to get it.

If you like commodities or not is not the point. The analysis that should be done is does its addition help or hurt a portfolio as a whole.

Good luck.
I do not think he is making that error.

To say beware of assets that do not make a return in and of themselves, is not contrary to Markowitz per se.

Implicitly, it usually includes the fact that to merely depend on correlations, going forward, to generate portfolio return is more iffy, than depending on (correlation + individual asset return). At a certain level, which I think the OP is certainly at, this is trivial.

If ones asset class, has ZERO expected return, you are putting it all on correlation for the portfolio. If the correlation benefit does not appear going forward, in ones time frame, one is left with not much benefit.

Wheres if the asset class has actual RETURN, then that benefits the portfolio. Its better to have both, expected return, and expected correlation benefit.

Correlations going forward are uncertain, so are returns, but I would rather have something that has BOTH possibilities for an asset class. If CCFs did have both, it would be a much better case for CCFs. Correlations and correlation benefit within a given portfolio are complex, and are often times very different from portfolio to portfolio. Simple Return of an asset class, not so much so.

LH

Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: Commodities: Positive Returns = Inconclusive

Post by Bradley » Sat May 05, 2012 1:03 pm

“It gets dug out of the ground in Africa, or someplace. Then we melt it down, dig another hole, bury it again and pay people to stand around guarding it. It has no utility. Anyone watching from Mars would be scratching their head.” :oops:
– Warren Buffett
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 1:31 pm

Bongleur
IN 2008 actually used correctly commodities certainly did no damage and likely helped. CCF allocation should be taken from the equity side because of the high volatility. The purpose is insurance remember so you don't want to be taking it from the bond side which would increase volatility, not dampen it. CCF did about as poorly as stocks. But investors in CCF should also have considered extending bond maturity because of the negative correlation to bonds (not a single year when both long bonds and CCF down). And that would have helped the overall portfolio some. Also keep in mind that the period is exactly when you would expect CCF to do poorly--in deflationary recession, when stocks do poorly and bonds do well. So again, why one should consider the whole portfolio. Same things happened btw in 81 and 2001. So nothing unusual. Note DFA now has an even cheaper version than PIMCO and doesn't actively manage the collateral as does PIMCO, but does actively manage the rolls.

LBill, the volatility of CCF and gold are way too high to be good inflation hedge in short term like quarterly or annually. As I showed the correlation for CCF and inflation rises over time, as one would expect as commodities are component of inflation, an input in production. Gold is not, There really is no logical reason for gold to hedge inflation. Just a belief. There is no very long period where CCF went way down and inflation went way up like there is with gold. Gold lost about 70% of its NOMINAL value while inflation was rising at about 4% a year for 24 years. Which is why IMO it is just wrong to consider it a good inflation hedge. Though I agree it has provided hedges against other risks, but not inflation.

Best wishes
Larry

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Sat May 05, 2012 2:07 pm

LBill, the volatility of CCF and gold are way too high to be good inflation hedge in short term like quarterly or annually. As I showed the correlation for CCF and inflation rises over time, as one would expect as commodities are component of inflation, an input in production. Gold is not, There really is no logical reason for gold to hedge inflation. Just a belief. There is no very long period where CCF went way down and inflation went way up like there is with gold. Gold lost about 70% of its NOMINAL value while inflation was rising at about 4% a year for 24 years. Which is why IMO it is just wrong to consider it a good inflation hedge. Though I agree it has provided hedges against other risks, but not inflation.
Yeh, I know it seems more logical for commodities to be a better inflation hedge than gold - input to production and all that. But if we call them Asset A and Asset B and just look at the numbers, I can't see it. From 1972-2011 the annual real returns for gold had a .37 correlation to CPI; whereas DJ-AIG real returns correlated .06, non-significant. We see that commodities real return had a negative correlation to inflation in 1972-1980 period (high inflation) and a non-significant correlation in the 1981-2000 deflationary period. Gold real returns were highly correlated with inflation in 1970s and negatively correlated for 1981-2000 deflationary period. But you didn't need an inflation hedge during 1981-2000 if you owned stocks and bonds, because both experienced large offsetting gains due to declining real rates and improving economy. You did need an inflation hedge in the 1970s, and that's when gold had outstanding returns. You needed a hedge against credit expansion, money printing, and negative real interest rates in the 2000s and once again gold has been a good hedge for that. In 2008, commodities lost -43% real, while gold gained 5% real. Gold is a highly liquid asset, while commodity CCFs locked up tight as a drum during the 2008 credit crisis. I pay my hedges to show up when they're needed the most, and I fire the ones that don't.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
Rick Ferri
Posts: 8594
Joined: Mon Feb 26, 2007 11:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by Rick Ferri » Sat May 05, 2012 2:21 pm

larryswedroe wrote:Unfortunately the statement you cannot eat lower volatility is incorrect. That is only true in isolation. Not in portfolio context. Lower volatility takes annual average returns and improves compound returns of a portfolio. You get that from assets with negative to very low correlation and high volatility. Best wishes Larry
Well, that's the HOPE is anyway. This ASSUMES that commodity prices will go up when stocks go down, which we know from 2008 isn't always that case. That being said, a small abount in commodities may not hurt the portfolio very much, I just don't beleive it helps the portfolio net-net in the long-term.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 2:29 pm

Lbill
As I said, IMO annual correlations meaningless basically with such high volatility. Need to look at annual data, and a 24 year period where you lose 70% nominal return and inflation averages 4% CANNOT be a good inflation hedge. Just cannot. So we will agree to disagree. Again, not arguing against including an allocation to gold, but not for inflation reasons

Rick
Well, that's the HOPE is anyway. This ASSUMES that commodity prices will go up when stocks go down, which we know from 2008 isn't always that case. That being said, a small abount in commodities may not hurt the portfolio very much, I just don't beleive it helps the portfolio net-net in the long-term.
As we have discussed many times, this is falsely states the argument for considering commodities. It is very logical that CCF should do poorly in deflationary recessions. Demand for everything falls. So CCF should fall. That is one reason that you take the allocation from equities. So likely it doesn't hurt any if at all because the stocks going down at same time, and you get the diversification benefits all the rest of the time. It turns out that about 70% of the time stocks go down CCF goes up and tendency is to be way up. That is pretty good diversification benefit IMO. But that is not the only consideration. 100% of time when bonds go down CCF goes up and tends to go up a lot. And not one single period when all three down. So the three have to be used, or should be used in connection with each other. In other words if adding CCF also add duration risk, and that works very nicely, hedging each other's risks well. Again, have to look at whole portfolio. The negative correlation of CCF to bonds is very logical. Bonds do poorly either because of inflation rising (when CCF does well) or real economy is stronger than expected, when demand for CCF is rising so it tends to do well. No reason to believe that won't continue. We do know that we should expect that there will be periods when stocks and commodities have negative correlation (supply shocks) and some when it will be positive (demand shocks). So IMO there is no hope needed, just understanding of basic economics.

Best wishes
Larry

User avatar
Rick Ferri
Posts: 8594
Joined: Mon Feb 26, 2007 11:40 am
Location: Georgetown, TX. Twitter: @Rick_Ferri
Contact:

Re: Commodities: Positive Returns = Inconclusive

Post by Rick Ferri » Sat May 05, 2012 2:42 pm

It turns out that about 70% of the time stocks go down CCF goes up and tendency is to be way up.
I know this is your hope. You've made that clear. My belief is that relying on a diversification effect to make up for no-real return from commodities, no income, and higher cost is a lot to hope for. But, if that's what you believe, good luck with it.

Rick Ferri
The Education of an Index Investor: born in darkness, finds indexing enlightenment, overcomplicates everything, embraces simplicity.

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Sat May 05, 2012 3:27 pm

When there was high inflation in the 1970s gold was a much better portfolio hedge than commodities. For example, the real compound annual return of gold from 1972-1980 was 22.5%, and for commodities it was 8.8%. An investment divided between the S&P500 and commodities or gold required twice the allocation to commodities as to gold in order to neutralize stock losses and produce a 0% real return over that period (7% vs. 3%). Likewise, during the period of 2001-2011 an investment divided between the S&P500 and gold or commodities required twice the allocation to commodities as to gold to neutralize stock returns and produce a 0% real return.

I find this interesting because it suggests that, when you needed real asset diversification the most, you only needed to carry about half as much in gold as you did in commodities to get the same bang. Bearing this in mind, had I entered the 1981-2000 period with half as much allocated to gold compared to a commodity allocation, my compound annual real return over this period would have been virtually the same. In fact, the half-gold allocation actually would have done fractionally better. For example, a 7% allocation to commodities (93% S&P500) resulted in a compound annual return of 11.0% with SD=13.44; whereas, a 3% allocation to gold (97% S&P500) resulted in an annual return of 11.3% with SD=13.9%. The Sharpe ratios were .66 and .65 respectively. Similar results with a 12% allocation to commodities vs. a 6% allocation to gold (10.8% vs. 11.1% with Sharpe ratios of .66).

This suggests to me that a smaller allocation to gold vs. commodities has the same "insurance" benefit with the same downside portfolio return risk. For example, if I were contemplating a 10% allocation to commodities I might be as well off from a hedging perspective with 5% in gold instead. If I were contemplating a larger allocation, such as 20%, to commodities then perhaps 10% to gold would do the job just as well. The "inflation hedging" argument is a red herring, IMO. You have to look at how these assets have performed in different time periods and what percentage allocation seems to be required to provide the left tail insurance that you need when you need it.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

User avatar
Lbill
Posts: 4997
Joined: Thu Mar 13, 2008 11:25 pm
Location: Somewhere between Up and Down

Re: Commodities: Positive Returns = Inconclusive

Post by Lbill » Sat May 05, 2012 4:29 pm

It's true that nobody can guarantee that an allocation to commodities or gold will deliver when it's supposed to. In addition to not investing in commodities, one could decide to not invest dead money into auto or home insurance and invest that money in stocks instead. After all, how often does your house burn down? And sometimes insurance companies don't pay claims when they're supposed to anyway. People who had hurricane insurance in New Orleans didn't get paid because the insurance companies said their house were demolished by flooding and not the hurricane, and they didn't have flood insurance. You just never know. But let's say that I invested $100,000 in 1972 and put 5% ($5,000) in gold, and split the rest between the S&P and Total Bond. Flash forward to 1980 and my $100,000 would have shrunk to $88,300 (real value) without the gold, but it was worth $101,810 with the gold, a difference of $13,510. So my $5,000 gold insurance premium kept me whole and produced a real return of 270% over that 9-year time period. If your number does come up, you must be present to win.
"Life can only be understood backward; but it must be lived forward." ~ Søren Kierkegaard | | "You can't connect the dots looking forward; but only by looking backwards." ~ Steve Jobs

larryswedroe
Posts: 15711
Joined: Thu Feb 22, 2007 8:28 am
Location: St Louis MO

Re: Commodities: Positive Returns = Inconclusive

Post by larryswedroe » Sat May 05, 2012 4:44 pm

I know this is your hope. You've made that clear
That is not correct. Those are the facts, which is quite different. IMO this is no different than hoping equities have a risk premium, which is based on historical evidence and logic. Same here. Again ignoring it is IMO the greater risk. But to each his own.
Finally I would add this, those with more exposure to the risks of unexpected inflation should give a stronger consideration to CCF, those with higher equity allocations should do the same, and those with more nominal bonds and longer duration should also do the same. Those with less have less need to do so.
Best wishes
Larry

Post Reply