How Commodities Can Help a Portfolio

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
User avatar
camontgo
Posts: 329
Joined: Tue Oct 27, 2009 1:46 pm
Location: Chicago, IL
Contact:

Re: How Commodities Can Help a Portfolio

Post by camontgo »

I think CCFs certainly provide some "insurance" benefit in the sense that they are likely to do well in some states of the world where other major asset classes are likely to do poorly.

The problem is that I'm uncertain about the expected cost of this insurance. Should we expect a positive risk premium for CCFs? I find it logical that stocks have a positive risk premium, but if an asset class, such as CCFs, is providing insurance against a widely held risk then shouldn't we expect the premium to be negative?

The historical return data for commodity futures looks attractive. As shown in the OP's article, there certainly appears to have been a large positive premium historically. This positive premium can also be seen longer term studies such as frequently cited Gorton and Rouwenhorst paper. However, the Gorton and Rouwenhorst reasoning for the cause of the historical premium is the Keynesian theory of "normal backwardation", which may have made sense in Keynes's era...but doesn't seem consistent with how people think about risk premiums in a world with modern asset pricing theory and integrated capital markets. In today's theory, you get a premium for holding (systematic) risk...not for buying insurance.

I'm very skeptical that the expected return on CCFs is positive (not counting the return on the collateral)...so I wonder if the insurance is worth the cost? Maybe it is, but too many advocates seem to place too much weight on historical returns.

I'd be interested to know what the expected risk premium that Larry and others would use as an estimate for CCFs going forward?
"Essentially, all models are wrong, but some are useful." - George E. P Box
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

camontgo wrote: As shown in the OP's article, there certainly appears to have been a large positive premium historically. This positive premium can also be seen longer term studies such as frequently cited Gorton and Rouwenhorst paper. However, the Gorton and Rouwenhorst reasoning for the cause of the historical premium is the Keynesian theory of "normal backwardation", which may have made sense in Keynes's era...but doesn't seem consistent with how people think about risk premiums in a world with modern asset pricing theory and integrated capital markets.

The OP’s article used the GSCI index and not an actual CCF. What’s wrong with the GSCI index? Below is a reference and a couple of snippets from the article they come from.

http://www.efficientfrontier.com/ef/0adhoc/stuff.htm


How can you not own these things? Easy. The planet described by Gorton and Rouwenhorst is not the one you and I live on.................

The Keynesian theory of normal backwardation . . . may fit the context of undiversified farmers during the 1930s, but it has less appeal in the context of modern multinational companies operating in global capital markets . . .

In other words, the next time someone tries to sell you a commodities fund based on the Goldman Sachs Commodities Index, smile and say, "Sorry, but I’m from Earth and you’re from planet I Love Lucy. Let’s revisit this discussion in an alternate universe."
-----------------William Bernstein


Any one who has read Bill's Four Pillars of Investing and The Intelligent Asset Allocator would be hard pressed to make the argument that he does not fully understand how commodities/CCFs work within a portfolio.

Bradley
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

few thoughts
first you can tell if something provides insurance by looking at the correlations of the returns with the type of risks
So it's clear from the evidence that CCF correlate with inflation positively, and the longer the term the higher the correlation, and more importantly they perform best in rising inflation

They also correlate positively with supply shocks (see 73-4) but negatively with demand shocks as the tables in my books and articles have shown (see 81 and 01),. So clearly they hedge some type of risks but not all risks, like most insurance--you buy the insurance for the risks you are trying to hedge/protect against. You want demand shock insurance buy LT safe bonds--credit risk safe

second re the future returns--I assume basically zero as there is no logical reason IMO to expect either backwardation or contango to continue. That leaves you with collateral return minus costs to get a small real return if managing collateral well and also perhaps from managing term structure well--buying cheapest months. If you get zero real return that would be good because you should expect somewhat negative real return with any insurance. And be willing to pay it as well

Best wishes
Larry
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: How Commodities Can Help a Portfolio

Post by rmelvey »

Hi Larry,

I just did a little messing around in simba's spreadsheet and I would appreciate your thoughts... I looked at excess returns (above the T-Bill return) for stocks, commodities, gold, and inflation.

With respect to inflation (in excess of the T-Bill rate) gold has a beta of 5.4, commodities have a beta of 1.06. With respect to the stock market, gold has a beta of -0.32 whereas commodities have a beta of 0.14.

To me this indicates that gold has a stronger allergic reaction to financial repression than commodities, as well as an additional hedging property with respect to equities (perhaps because of its semi-monetary properties). What are your thoughts on this?

To me it says that gold gives me more "bang per buck" and the negative correlation means that I can hold more equities than I could if I was using CCF.
muntz
Posts: 84
Joined: Mon Apr 14, 2008 5:33 am

Re: How Commodities Can Help a Portfolio

Post by muntz »

Below are the thoughts of some of the brightest/most experienced minds in money management today which express their opinions on including commodities in a portfolio. Before learning the hard way, anyone thinking about adding commodities AND then adding term risk should first consider the following comments on commodities. I'm quite sure these men are very familiar with Modern Portfolio Theory.
ok...now we've taken a turn into different territory. we are not discussing whether they belong ina portfolio. that is an entirely different subj. you were having trouble understanding the idea of insurance.

Also, Larry is an expert as well.
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

muntz wrote:Bradley - Commodities protect agst certain types of inflation risks that could harm the rest of your portfolio.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

rmelvey
gold as I explained hedges loose monetary policy and the risks associated with it --not inflation itself
Larry
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: How Commodities Can Help a Portfolio

Post by rmelvey »

Hi Larry,

Since monetary policy has such a huge affect on the bond market (explicitly altering the short end of the curve, and implitly changing the long end because of pure expectations hypothesis), couldn't one make the argument that gold is the better inflation hedge in the context of a diversified stock and bond portfolio. After all, in this context we only really care about the inflation that is not compensated for in bond yields. The inflation of the 1980s was not really painful at all because nominal yields more than made up for it.
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

larryswedroe wrote: So it's clear from the evidence that CCF correlate with inflation positively, and the longer the term the higher the correlation, and more importantly they perform best in rising inflation

Larry
Not so sure the evidence is as clear as you claim it is.



http://online.wsj.com/article/SB1000142 ... 45096.html

February 5, 2013, 9:08 p.m. ET

Pension Funds Cut Back On Commodity Indexes

"Pension funds and other institutions are retreating from popular investments linked to commodities after finding they did little to protect their portfolios against inflation risk and the unpredictable returns of stocks.....Investors have yanked nearly $10 billion from tradable indexes tied to energy, food, metals and other commodities after two years of record outflows......The trend is accelerating this year, analysts and investors said, driven by lackluster returns and looming U.S. regulations that could make these investments more complicated and costly.......Among those scaling back is the California Public Employees' Retirement System. Calpers, the nation's largest pension fund, pulled out 55% of its holdings in commodities indexes in October, after losing about 8% annually over five years, according to the fund's most recent financial statement."


Please do not claim CALPERS, Bogle, Ferri are considering commodities/CCFs in isolation. IF they actually did help portfolio performance pension funds/institutions would not be rethinking the use of commodities.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Roy
Posts: 970
Joined: Wed Sep 10, 2008 9:34 am

Re: How Commodities Can Help a Portfolio

Post by Roy »

Bradley wrote:IF they actually did help portfolio performance pension funds/institutions would not be rethinking the use of commodities.
I don't own commodities. But I'm not willing to give these guys a pass; and this is not evidence that commodities don't help.

Active managers like these and many others (401ks etc.) chase asset classes that are currently working best and abandon what isn't. The reason is they must constantly outperform or get replaced, and so style shifting is rampant. It doesn't work, and so the managers tend to be replaced frequently. And that applies to Small Caps, High Yield, International, and other asset classes they chase when fashionable.
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

Roy wrote: this is not evidence that commodities don't help.

If a 40% loss over five years in commodities is not evidence that the asset class is not working what would be?

CaLPERS has more indexed funds under management than any other pension fund and with near $250 billion in assets. I think it is fair to say they can afford the best advice available.


............ Illinois Teachers’ hired three groups in February 2008 to manage its first commodities investment of $600m, announcing the fund was taking “additional steps to protect retirement nest eggs”. By June 30 2008, the investment had swelled to $652.8m. Two days later, what was then known as the Dow Jones-AIG commodity index peaked before tumbling 57 per cent in the financial crisis.
In June 2011 the Illinois commodities portfolio was worth $324.7m, according to a fund report. In May this year the fund eliminated its commodities allocation ........


Can't help but think commodities do not help.

Bradley


Jack Bogle's advice on commodities in portfolio?


Should not be in “anybody’s portfolio, at anytime, under any circumstances. Did I make that clear?”
----------------------------------Jack Bogle
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

Bradley
When I present a case I present the evidence to support the statement.
Like I have with the statements about the correlations, the data has been presented here by me many times and in my books, those are facts and not debatable issues. Just because they don't fit your story doesn't change the facts
Now if you want to present facts/data to back up your statements like you're not sure it's as clear as I say then please do so. your statements without facts or data have no value. So where are your facts?

As to your statements re "experts" --what matters are facts, not opinions. And as perfect example you cite Bogle. While I have the greatest respect for him John has clear been wrong about major issues in the past. Like don't need to diversify internationally and also statement in his telltale chart that value stocks had not outperformed growth stocks. Clearly the first isn't even debated by any academic I know of and the second is wrong on its face---making the mistake of using mutual fund, and active funds at that, instead of looking at the stock performance. Now you if you want to provide some real evidence Bogle presents rather than just his opinions, then we can discuss those facts.


As Roy stated pension plans are among the worst investors and are known performance chasers, citing them as experts doesn't help your case at all.


so we all wait for your facts using portfolios, not looking at things in isolation to make your case. Not opinions but facts/data
Larry
Larry
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

larryswedroe wrote:Bradley
When I present a case I present the evidence to support the statement.

Now you if you want to provide some real evidence Bogle presents rather than just his opinions, then we can discuss those facts.


Larry

Larry,

The evidence you cite in your article to advance your theory on commodities is silly. Data should be objective and not biased toward the “researcher’s” theory. For example, your use of an index vs a fund which bolsters your portfolio’s performance. You show no cost involved. You use inappropriate benchmarks.

When you use an index vs a fund you first give your self the advantage of not accounting for the funds e/r. If you used an investable fund like PCRIX, your recommended fund, you would have had to add 74 basis points per year to the cost of implementing your strategy. 74 basis points over the 42 year period you use in your article adds up. This alone makes your “facts” very questionable.

In “your portfolios” you use the S&P 500 vs more appropriate benchmarks. Once again this cherry picking slants results in favor of your theory. Do you recommend the S&P for your clients? Then why use it as a benchmark. This is what active managers do to enhance their performance. An objective study would have used Vanguards Total Market or one of DFA’s core domestic equity funds and not the underperforming S&P as you do.


Using the GSCI in any research is questionable especially when using that data to recommend a fund(PCRIX) which attempts to track the Dow Jones UBS Commodity Index. Here is what Bill Bernstein has to say about the index you use to sell your strategy,............”the next time someone tries to sell you a commodities fund based on the Goldman Sachs Commodities Index, smile and say, "Sorry, but I’m from Earth and you’re from planet I Love Lucy. Let’s revisit this discussion in an alternate universe."



Every step of the data presented leans in favor of the outcome you attempt to prove.

Come back using an investable commodity fund, appropriate “portfolio funds” and up to data data then we’ll talk. Otherwise Mr Bogle's opinion stands.


Bradley
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

Bradley
I won't waste any more time with your nonsensical comments
There is nothing silly at all about the returns data I have shown. Data even the same for global portfolios and value tilted portfolios
If I showed them as well you'd just make up some other stuff. so I did not waste the time. But it's the same.

In fact the I've shown clearly that the benchmark indices likely UNDERSTATE the returns a well run fund has been able to earn
Also I showed the data for live funds anyway, And for the period that has been one that CCF hedges were not needed-=-=no supply shocks and no unexpected inflation and yet despite your claims and Rick's have shown that even in this period adding CCF did slightly improve returns, and this was one during which there was persistent contango as well--something that is now gone as I basically predicted because if you get persistent contango it impacts returns and thus demand for CCF will fall. It is exactly what has happened.

As to S&P I already explained why your comments were just plain wrong. First the S&P is basically the market, about 80% and a very common benchmark and over long term return between it and market as whole is almost the same. And if you ran the data with other benchmarks like TSM or global economy you get the same results. I've shown that in my books. You just don't like it
I've never cherry picked data and in fact done exactly the opposite, choosing periods when CCF did poorly relatively so to show how it performed during such periods--like 2007, refuting Rick's statement that those who chose to add CCF did poorly since then, That is simply wrong, especially if used longer bonds as I advised

Now you persist with no facts and when shown the facts you never acknowledge them, so bye bye
Just fell sorry for those that take your statements seriously

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

Re: How Commodities Can Help a Portfolio

Post by Bradley »

larryswedroe wrote: There is nothing silly at all about the returns data I have shown.
Larry


Larry,

This is what William Bernstein had to say about commodities and the GSCI upon which you build your case.



“Why don't you ever write about commodities?

Because I don't trust the data, particularly the GCSI index, which reportedly has a real return of 5% and zero correlation with the rest of the market over the past three decades. This makes no sense at all during a period when the underlying commodities have had a negative real return. It's just too good to be true..........................................................................In other words, the next time someone tries to sell you a commodities fund based on the Goldman Sachs Commodities Index, smile and say, "Sorry, but Iʼm from Earth and youʼre from planet I Love Lucy. Letʼs revisit this discussion in an alternate universe."
-------------------William Bernstein





There are no live funds in your article,none. You rely on GSCI index returns which probably explains why Bogle, Ferri, Bernstein, French and Fama..........etc do not agree with your theory. Garbage in, garbage out. Your post/explanation would no doubt sway many non sophisticated investors but most forum members understand how the real world has cost associated with your theory which you ignore in your article. Most members also understand how using the S&P vs Total Market or DFA core fund weakens the credibility of your "proof". To infer that a funds e/r is inconsequential to the outcome of tracking an index is wrong and that alone renders your “data” unreliable.




Bradley



John Bogle: I for one--I hope it's all right for me to say this--have no conviction that commodities belong in anybody's portfolio, at any time, under any circumstances. Did I make that clear?
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

Sorry, I forgot to include the opinions of Ken French and Gene Fama, long time opponents of commodities in most investor portfolios over the years. If these guys do not understand how commodity futures affect one’s portfolio, who does?



Nov 8th 2008

So is this a once-in-a-lifetime opportunity to climb about the commodities train, as many burned advisors are now saying?

No, says Ken French, professor of finance at Dartmouth's Tuck School and the director of investment strategy at Dimensional Fund Advisors. There's no reason for most investors to own commodities. Contrary to popular belief, they aren't a good inflation hedge, and they don't provide a long-term real return that investors aren't already exposed to through normal stock ownership.”




--------------------------------------------------------------------------------------------------------------------
Below is from the French/Fama page on DFA's website.

The claims that, going forward, commodity funds (i) will have the same Sharpe ratio as the stock market, (ii) will be negatively correlated with the returns on stocks and bonds, and (iii) will be a good hedge against inflation can't all be true. Who would want the other side of this trade? The high volatility of commodity prices makes it impossible to accurately estimate the expected returns, volatilities, and covariances of commodity funds, but theory suggests that if commodity returns are negatively correlated with the rest of the market, the expected risk premium on commodities is small, perhaps negative. Finally, commodity funds are poor inflation hedges. Most of the variation in commodity prices is unrelated to inflation. In fact, commodity indices are typically 10 to 15 times more volatile than inflation. As a result, investors who use commodity funds to hedge inflation almost certainly increase the risk of their portfolios. Six years after my presentation, I still think these points are correct.
-----------------------Kenneth R. French........................He is the Roth Family Distinguished Professor of Finance at the Tuck School of Business at Dartmouth College. He is an expert on the behavior of security prices and investment strategies.

Professor French is a consultant to Dimensional Fund Advisors and a member of the firm's board of directors. He is also a regular speaker at Dimensional conferences and seminars.
French is a research associate at the National Bureau of Economic Research, an advisory editor of the Journal of Financial Economics, a former associate editor of the Journal of Finance and the Review of Financial Studies, and a former president of the American Finance Association. French is also a Fellow of the American Finance Association and the American Academy of Arts and Sciences, and a member of the Smile Train's board of governors and the International Rescue Committee's board of directors.
Before joining Dartmouth, Professor French was on the faculty of MIT's Sloan School of Management, the Yale School of Management, and the University of Chicago Booth School of Business. Professor French received his PhD in finance from the University of Rochester in 1983. He also earned an MS and an MBA from the University of Rochester and a BS from Lehigh University.






Bradley


"You're picking up nickels in front of a steamroller," says William Bernstein, an author on the subject of the dangers of investing in commodities.
Last edited by Bradley on Sun Jun 09, 2013 7:25 am, edited 1 time in total.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

There have been commodities funds around long enough that one could do a simulation based on actual returns, rather than relying on an index:
Go back as far as real world funds permit
Use TSM and TBM as the stock and bond investments, covering the US markets.
Create a portfolio, updated monthly, of all the commodities funds available at the time. If funds close or are merged out of existence, keep their records in the data up to the point that they close. No survivorship bias
Combine the commodities funds into an equal-weighted portfolio and a cap-weighted portfolio. They address different questions

Compare an overall portfolio of TSM with TBM a portfolio of TSM, TBM, plus equal or cap weighted commodities funds. Vary the weights of stocks, bonds, and commodities.

Compute monthly returns and generate the standard statistics, mean, variance, single, 3 and 4 factor beta and alpha. To do a fair comparison, one might have to include a 5th factor that would attempt to capture the return of commodities through multiple indexes. Picking the "best" one might be impossible, and it would be more informative to use several. That would avoid attributing to "good management" the success of a commodities fund that got its performance from something other than commodities.

The results will compare real world performance of stocks plus bonds to the performance of stocks and bonds plus commodities.

Stop arguing about what the results would be until we see them.

I don't have the energy to do this, since I am currently not interested in investing in commodities. But there seem to be enough people who are engaged by the issue that I hope someone will put out the effort to do the study.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
technovelist
Posts: 3611
Joined: Wed Dec 30, 2009 8:02 pm

Re: How Commodities Can Help a Portfolio

Post by technovelist »

Is gold a commodity? There is indeed evidence that incorporating gold in one's stock/bond portfolio reduces volatility while not hurting return to any significant degree. See http://www.bogleheads.org/forum/viewtop ... 0&t=116025 for a discussion of this.
In theory, theory and practice are identical. In practice, they often differ.
wesleymouch
Posts: 237
Joined: Wed Dec 05, 2012 1:24 pm

Re: How Commodities Can Help a Portfolio

Post by wesleymouch »

The anti commodity crowd has an air of religious intolerance about it. Accept that there can be different views of the world and investing in it. No one is forcing you to invest in anything. Get over it.
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

I think this chart clearly shows that adding commodities would have helped if rebalancing bands were used in the period from 2008 on.

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

not sure how much only 10% commodities would help
Last edited by TO39 on Thu Jun 06, 2013 4:01 pm, edited 1 time in total.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

It is not obvious to me from the chart. It may be that there would be a reduction in variance without much penalty, or perhaps an increase, in mean return. The chart could not tell us how well the commodities fund tracked the commodities index, even assuming one knew which index to tract. You would really need the variance, mean, alphas and betas.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

To AFAN

It is not obvious to me from the chart. It may be that there would be a reduction in variance without much penalty, or perhaps an increase, in mean return.
That sounds to me like it helped. lower volatility without much cost, just like Larry said


The chart could not tell us how well the commodities fund tracked the commodities index, even assuming one knew which index to tract. You would really need the variance, mean, alphas and betas
Some one requested what would a real fund do vice an index, so I provided one with the most history I could find.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

TO39,

I appreciate your efforts, but as I said, I cannot tell from the chart whether the combination of commodities with stock and bond funds produced a lower variance than the stock and bond funds alone. Maybe, maybe not. One would need to see the actual numbers.

These are results from one fund, but cannot be assumed to be representative of commodities funds, do not account for survivorship bias, and as I mentioned, it is not obvious which commodities index one would want to follow. So we can see that the commodity fund did not have a perfect correlation with stocks or bonds, but I don't know whether the correlation was negative or zero. If I had to guess, I would think it was positive. Again, one would need to see the numbers.

And the concerns about picking only a fund that did well enough to survive and be prominent obtain. To test commodities as they would have been invested over time, one would have to look at the combined performance of all the commodity funds that were available for each time period. I don't know whether you could get that from Morningstar for funds that closed or merged, I never tried. It would not be trivial, which is why I have not done it, but it could be done by someone who really wanted to know how well commodities actually worked to improve mean/variance performance over time.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

To afan

My understanding is DBC is an index fund that tries to replicate the commodities market, not beat it. there would be tracking error, but that should balance out to zero. And again someone requested what did a real commodities fund with expense ratio do. This was the best I could find.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

Right. It is one fund, that tracks one index. We don't know whether one would have picked this particular index back when the fund started, we don't know how other funds that sought to track similar indices fared, or whether this index made sense. This fund underperformed the Morningstar "best fit" index, but again, who knows whether that was the right index either? The index this fund attempts to track was not an obvious one, since it is not cap weighted. It is a mix of various commodities, some weighted far more than their market representation, some much less. Change the mix of weights, and you change the performance. But what should the mix be? What would the mix have been for an investor deciding to "put some money into commodities" back when the fund started? Some of the returns came from holdings of Treasury securities, so the fund performance was not a pure commodities exposure.
he PowerShares DB Commodity Index Tracking Fund seeks to track changes, whether positive or negative, in the level of the DBIQ Optimum Yield Diversified Commodity Index Excess Return™ ("DBC Commodity Index" or the "Index") plus the interest income from the Fund's holdings of U.S. Treasury securities less the Fund's expenses.
It is somewhat interesting to see the performance of one fund, but it does not address Larry's point about viewing commodities in the context of an overall portfolio. For the reasons in my posts, this does not tell us what a real world investor who decided to add commodities would have experienced on the overall portfolio, even if she/he happened to pick this fund.

It is not a massive undertaking to figure this out, but there is no way the performance of one fund, selected at the end of the period, could answer the questions people have raised about Larry's observations.

I am not interested in commodities funds, in part because I am not convinced that the benefits can be realized in the real world ex ante, and in part because I worry that people are still figuring out how to run them without dramatic underperformance due to the factors others have raised. Plus, many of them avoid some real world investment problems by being ETN's, but that introduces the risk of the creditworthiness of the issuer of the ETN. It is difficult to account for that risk by looking at standard deviation, but it is a big problem if the company goes under.

With those concerns, I will stay on the sidelines until people like Fama have sorted it out. I know he is quoted as unenthusiastic about commodities as an investment, but as far as I know he has not produced a systematic study of returns in the context of a combined stock/bond/commodities portfolio.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
User avatar
Ketawa
Posts: 2521
Joined: Mon Aug 22, 2011 1:11 am
Location: DC

Re: How Commodities Can Help a Portfolio

Post by Ketawa »

Seems like some people have an axe to grind and just resort to arguments from authority. This is not very convincing against arguments with a good theoretical explanation and actual data, especially when the authorities in question have given dubious advice on other investing questions. Maybe this article didn't include any real world results, but Larry has posted the results of including commodities funds in a diversified portfolio in other threads. See this link.
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

Ketawa wrote:Seems like some people have an axe to grind and just resort to arguments from authority. This is not very convincing against arguments with a good theoretical explanation and actual data, especially when the authorities in question have given dubious advice on other investing questions. Maybe this article didn't include any real world results, but Larry has posted the results of including commodities funds in a diversified portfolio in other threads. See this link.

agree with this, except the authority part, maybe they just argue to hear themselves. repeated assertion that another fund would do this or that, but never the charts or data that back it up
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

Not sure if that was addressed at me, but my point is that, just as one can always identify a stock fund that did better than the index, one can also find a commodities fund that did well. Since the funds vary in their mix and weights of assets, a good performance might be due to chance that the fund happened to be in winning commodities. Since no one can know what these will be going forward, a meaningful analysis would have to do what one does for stock funds- create portfolios of all eligible funds, without requiring that the funds survive the entire period. With so much diversity in commodities indexes, one cannot even choose an index fund and assume this is representative of the market as a whole.

There is now agreement that the stock market as a whole can be modeled as a cap weighted portfolio. one can determine factor contributions, but there is not a lot of controversy over the cap weighted index. For commodities, many people find that the cap weighted indexes and "too" heavily weighted in energy, so they pick something else. Once you do this, you have stopped passively investing in the "commodities market" and are instead making a call on which commodities to over and underweight. So pick a fund that uses some index as its benchmark, but then does not actually track that index. Do this after the time period is over. How could one possibly relate the results to ex ante expectation?

An illustration that uses a single commodities fund to represent the contribution of commodities as an asset class would be as meaningless as if one chose a single active stock fund to represent the stock market.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

TO39 wrote: repeated assertion that another fund would do this or that, but never the charts or data that back it up
Here is some data. The fund the OP uses and recommends for commodities is PIMCOs PCRIX. The fund went live in 2002 so you can look at ten year data. The Commodity fund returned $16,997.19 over 10 years compared to $21,473.95 for Vanguard’s Total Market Fund based on an intial $10,000 investment. That works out to an outperformance of nearly 3%/year for 10 years. Over the past five years the total market fund has outperformed the commodity fund in excess of 13%/yr on average over each of the past 5 years. All this data is available at M*.

Hope this helps,
Bradley
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

Bradley wrote:
TO39 wrote: repeated assertion that another fund would do this or that, but never the charts or data that back it up
Here is some data. The fund the OP uses and recommends for commodities is PIMCOs PCRIX. The fund went live in 2002 so you can look at ten year data. The Commodity fund returned $16,997.19 over 10 years compared to $21,473.95 for Vanguard’s Total Market Fund based on an intial $10,000 investment. That works out to an outperformance of nearly 3%/year for 10 years. Over the past five years the total market fund has outperformed the commodity fund in excess of 13%/yr on average over each of the past 5 years. All this data is available at M*.

Hope this helps,
Bradley

That does help. We're halfway there. What would the total portfolio return and deviation have been for those two periods, if you had annual rebalancing, 20% or 40% band rebalancing? I would like to see the comparison between 60/40 stock/bond portfolio and 50/40/10 stock/bond /pcrix portfolio.

Don't you agree that it is the total portfolio performance that matters and not the individual components?

http://quote.morningstar.com/fund/chart ... %2C0%22%7D


After looking at the above chart, I think it adds credence to the idea CCFs can help total portfolio performance.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

T039,

I am terrible at glancing at charts and determining alphas, betas, and variance. This means I could not begin to guess Sharpe ratios. I think I am a little bit better with guessing correlation coefficients, but not much. Here I would say that the commodities fund and the stock fund had a significant positive correlation, with no guess as to the value of the coefficient.

I have no idea how much of the return to the commodities fund came from the performance of the underlying commodities, versus trading strategy and bond returns. I have no idea whether the addition of commodities raised or lowered the Sharpe ratio.

If you can tell what these values are for the stock and bond portfolio and the stock, bond, and commodities portfolio, can you tell us?

Thanks

Afan
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
CashIsKing
Posts: 39
Joined: Thu Feb 24, 2011 9:15 am

Re: How Commodities Can Help a Portfolio

Post by CashIsKing »

nisiprius wrote:
larryswedroe wrote:... need to stop thinking about commodities as creating wealth the way stocks are expected to do--that's the problem. You need to think of them as portfolio insurance.
I do not think investments should be called "insurance" (and I do not think insurance should be called "investments.")


Insurance is a contract in which an accountable party promises to do specific things if certain things happen. Some vague historical tendency for some asset to behave in certain ways isn't insurance. If it doesn't do it what you expected it to do, whom do you take to court? Wouldn't you take issue if a gold advocate called gold "insurance?"

The specific phrase "portfolio insurance" is further confusing because when I hear it I think of the crash of 1987. The term has a specific technical meaning: it refers to some complicated hedging system (having nothing to do with commodities) that failed spectacularly. I'm sure you know orders of magnitude more about what it was and what it did than I do.
nisiprius,
Actually, your definition of insurance deals with just one specific dimension of the concept. The most fundamental and accepted definition is that with insurance you exchange the potential of a large loss with the certainty of a small(er) loss. IMO, that is precisely what Larry Swedroe is proposing. Whether one agrees or disagrees with the inclusion of commodities in their portfolio is a different discussion. But a specific component of a portfolio that generally, predictably, moves in the opposite direction of another component can rightly be called "insurance" because it does what insurance (see above) is supposed to do. It's ridiculous to think of it as anything else, because of the fact that there's no contract or the ability to take someone to court. It is not Auto Insurance or Homeowners Insurance; it is portfolio insurance (please note the lower case "i")
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

CashIsKing wrote: nisiprius,
....... The most fundamental and accepted definition is that with insurance you exchange the potential of a large loss with the certainty of a small(er) loss. .............. But a specific component of a portfolio that generally, predictably, moves in the opposite direction of another component can rightly be called "insurance" because it does what insurance (see above) is supposed to do. It's ridiculous to think of it as anything else, ............. it is portfolio insurance (please note the lower case "i")


My “insurance” policy is the TBM. I pay .10 for that “insurance”. The commodities fund recommended on this forum has been Pimco’s PCRIX. That “insurance” is 7.4 times as expensive as my “insurance”. In 2008 the market lost 36.98% as measured by the TM. My low cost, easy maintenance TBM returned 5.15%. The commodity “portfolio insurance” policy lost 43.33%. This would be a heart stopping event for many investors, especially those who count on paying the bill from their portfolios. Portfolio "insurance" that only works randomly is not a policy I would renew.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
CashIsKing
Posts: 39
Joined: Thu Feb 24, 2011 9:15 am

Re: How Commodities Can Help a Portfolio

Post by CashIsKing »

Bradley wrote:
CashIsKing wrote: nisiprius,
....... The most fundamental and accepted definition is that with insurance you exchange the potential of a large loss with the certainty of a small(er) loss. .............. But a specific component of a portfolio that generally, predictably, moves in the opposite direction of another component can rightly be called "insurance" because it does what insurance (see above) is supposed to do. It's ridiculous to think of it as anything else, ............. it is portfolio insurance (please note the lower case "i")


My “insurance” policy is the TBM. I pay .10 for that “insurance”. The commodities fund recommended on this forum has been Pimco’s PCRIX. That “insurance” is 7.4 times as expensive as my “insurance”. In 2008 the market lost 36.98% as measured by the TM. My low cost, easy maintenance TBM returned 5.15%. The commodity “portfolio insurance” policy lost 43.33%. This would be a heart stopping event for many investors, especially those who count on paying the bill from their portfolios. Portfolio "insurance" that only works randomly is not a policy I would renew.
Bradley,
Evidently you did not understand my post. I specifically pointed out that whether or not commodities are a wise or valid form of "portfolio insurance" was totally irrelevant to my point, which was about what "insurance" is at its most fundamental. Actually, your post supports what I said: You give up higher gains (or even losses) through TBM, to eliminate the potential of much larger losses via a 100% equity position.
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: How Commodities Can Help a Portfolio

Post by rmelvey »

I feel like most of the commodity detractors in this thread here haven't internalized the difference between a supply shock and a demand shock. Citing the 2008 crisis as a reason to not hold commodities is a prime exhibit of this.

If you hold the stock market:
-bonds offer a demand shock hedge
-commodities offer a supply shock hedge

Demand shocks have occurred numerous of times in the last 30 years, whereas the last great supply shock occurred in the 1970s. You can bring out all sorts of empirical data about how they haven't done what you would like in the past 20-30 years, but Larry is talking about them hedging against a theoretical event that is out of sample (unless you go back to the 1970s). Focusing on correlation coefficients is missing the big picture.

Asset markets respond to economics. There isn't some black box out there that plugs in a correlation and variance and spits out asset returns. Asset returns are driven by macroeconomic causations. You can use simple macro theory to understand the diversifying properties of certain investments, ignoring their ex-post correlation coefficients. The men quoted earlier in the thread are all respectable, but those are just sound bites meant to prevent retail investors from doing something they don't understand. I like to take personal responsibility to understand things for myself, and the theory and data says to me that commodities can play a role in certain portfolios.
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

rmelvey
In addition, almost every statement cited is either incorrect or irrelevant--I have written about every one of those issues many -times explaining why they are either wrong or irrelevant-One get's tired of having to repeat the same explanations

Just would add two things
first is that while longer term SAFE bonds hedge demand shocks, TBM doesn't do as good a job for the many reasons I have cited, and there is nothing to hedge the risks of unexpected inflation, either for stocks or bonds and it can hit both

Second, CCF provides a hedge against some risks as I have explained, and by doing so you cut both tails, not one. So when the risks don't show up you should not be surprised to see lower returns with CCF as the trade off for the higher (or less worse) returns when the risks do show up. People who don't understand this confuse strategy and outcome and thus make poor portfolio choices

When I get a chance I might write a piece explaining why in each case the French/ Fama an d other quotes are either wrong or irrelevant, and it is very easy to show it in each case.

Best wishes
Larry
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

rmelvey,

Can you do this for commodities? And explain how, if one ignores ex-post returns, one can be sure the macro theory was correct, and applied correctly?
use simple macro theory to understand the diversifying properties of certain investments, ignoring their ex-post correlation coefficients
Larry,

I have not read all your writing. I am just asking to treat commodities the way we would treat a study of stock investing. We don't pick a single fund to represent stocks, we look at the universe of stock fund. We don't pick a single fund, then compare it to an index that it does not track and congratulate it for "outperforming". We don't ignore all the other funds that existed during the period. If we pick an index, we provide a reason why that index was chosen. The index cited above does not seem to have an economic justification.

Do you know of studies done with proper attention to the universe of commodities funds, exclusion of biases, and compared realized performance of stock/bond portfolios with and without a contribution from the mix of commodities funds available for investment during the time?

Thanks
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

afan
IMO that is what has been done.
A) Prior to the 2002 intro of PIMCO fund, and even then if memory serves they did not lower their fee till later, there were no passive funds. so it's irrelevant to look at funds.
B) there are two major indices, the GSCI which is energy heavy and thus more subject to possible major contango, and the DJ UBS which is more diversified and thus not as subject--which are the two reasons it was developed.
C) While the PIMCO fund does cost 74 bp as I showed simple trading strategies and fund construction rules have allowed both PMCO and DFA to outperform their benchmarks by quite a bit---so looking at the index data likely is too conservative in what would have happened.


Other indices have been created since to avoid some of the problems of opaque indices and also contango.
that is just smart construction, just as using the CRSP indices was always better than using the R2k, which cost Vanguard investors about 2% a year for the privilege of avoiding tracking error until Sauter got the board to allow him to change

I hope that is helpful
Larry
User avatar
rmelvey
Posts: 826
Joined: Sat Sep 18, 2010 5:17 pm
Contact:

Re: How Commodities Can Help a Portfolio

Post by rmelvey »

afan wrote:rmelvey,

Can you do this for commodities? And explain how, if one ignores ex-post returns, one can be sure the macro theory was correct, and applied correctly?
use simple macro theory to understand the diversifying properties of certain investments, ignoring their ex-post correlation coefficients
Well, to my knowledge, we cannot witness historical demand and supply curves so we have to take a more qualitative look at characterizing certain time periods. The first step is to understand what a supply shock is.

Image

A supply shock occurs when the aggregate supply curve shifts left. As you can see in the image, it entails higher prices (manifesting itself in higher inflation), and lower quantity (manifesting itself in lower real GDP). The 1970s is a text book example of this. Oil supply to the US was dramatically reduced. This shifted the curve left. Throughout this period the economy had rising prices and bouts of declining real GDP. It makes intuitive sense as well. If we have less oil, it makes sense that prices for goods would be higher and less goods would be available to purchase. This is the unique risk factor that commodities hedge against. Bond holders would be hurt by the rising price environment, and equity holders wouldn't like the declining quantity of output environment. Commodities offer a "bird in the hand." Sometimes that is extremely valuable.

EDIT:

If one is truly concerned about the biased results of backtested indices, one can simply observe metals such as gold and silver. They have been perfectly investable in from the 1970s onward, and they tell a similar story to the commodity futures indices. It is possible that returns for the indices are over/understated, but I don't think that the diversification story has been manipulated at all. It's just basic macro reasoning.
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

amazingly enuff, the Morningstar growth chart now has pcrix outperforming VTI since inception. They added another 8 months to the starting period.

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

apparently there were some shocks in this time period also. unless I did something wrong in pulling up the chart
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

Try VTSAX the Total Market Mutual fund. It has beaten PRIX over YTD, Past yr, past 3yr, past 5yr and the past 10 years.
Last edited by Bradley on Fri Jun 07, 2013 4:24 pm, edited 1 time in total.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

That was my point. now Morningstar has 12 years of data for VTI and pcrix. They have both changed. Now PCRIX since inception out performs VTI and VTSMX

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

look for yourself, if I am doing something wrong, let me know what it is.



edit maybe morning star is adding more data
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

Thanks for pointing that out.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

Well, I don't know about "outperformed". PCRIX had a slightly higher mean return than VTI, but it looked much more volatile. I can't get SD or mean for the full period from Morningstar. For 10 years the means were VTI 8.35, PCRIX 5.41. SD VTI 15.16, PCRIX 21.50. Sharpe ratio VTI 0.50, PCRIX 0.28.

Larry,

On the one hand, I agree that it is problematic to look at "commodities" returns before there were any passive funds. However, if one looks at the entire commodities experience, using active commodities traders as the proxy, then one can see how the market performed. One would have to impute a set of costs for the active traders, but looking at this would still provide a look at the performance of commodities.

Since PCRIX does not actually track the UBS index, and it is not a passive fund, then it is not a good fund to tell one what would have happened if one had tracked the index. At best, if one wanted to use this index, one would still have to look at the index returns, subtract something for expenses, and use that to estimate what the commodities return would have been. If one wanted to also credit for the bond-like return included in the fund, that could be a positive for the portfolio, but it is not a return to commodities. One could also identify a trading strategy that took advantages of weaknesses in the construction of the index, and predict that one could beat this index by some amount, and credit that to a commodities return. However, then one has just wandered off indexing, and perhaps picked a "weak" index as a straw man benchmark.

There is still the problem that picking an index that intentionally is not representative of the commodities universe raises the usual concern that the performance of the index may not tell us what was the performance of commodities.

I seem to be one of the few people commenting on this who is certain that he does not know whether "commodity" investments would have had the desired effect, since there does not seem to be data that would answer the question. I agree it would be hard to get this data. But lack of data does not justify the level of confidence, positive and negative, expressed on this thread.

When there is no data, the logical answer would seem to be "we don't know, no one has the data", rather than "since there is no data, I will take a very limited sample, very different from the overall performance of commodities, and assume it must reflect what would have happened".

In other words, limited evidence should reduce certainty, rather than increase it.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Bradley
Posts: 486
Joined: Tue May 01, 2007 2:41 pm

Re: How Commodities Can Help a Portfolio

Post by Bradley »

afan wrote:


There is still the problem that picking an index that intentionally is not representative of the commodities universe raises the usual concern that the performance of the index may not tell us what was the performance of commodities.

The oldest commodity index data was published in 1958(CRB). It still trades on the exchange and like the other indexes has changed the way it is calculated. It was originally calculated by the Commodity Research Bureau. It’s energy weighting is about 1/3 of GSCI which started up in 1992.
You can sum up any active fund manager’s presentation at an investor conference in one sentence: “We’re doing well, all things considered.”
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

afan
I don't agree with your thinking here at all.
Looking at active funds tells you nothing about the performance of the passive strategy, so it's useless. They can be in and out of particular commodities and so on.
PCRIX is totally passive on commodities and eliminates or tries to minimize the negatives that pure indexing has, especially in commodities. It avoids the 5 roll dates the index rolls on and it uses the academic research on futures curves to find the cheapest months to roll. Now it is very active on the collateral side and has a good strategy IMO of using TIPS instead of tbills> Now why would you not want to do that. TIPS gives you a double real return asset and allows you to go out on the curve without inflation risk. That is the good side of their active strategy. The negative IMO (though they have added value) is that they add in riskier assets like EM debt and MBS to try and pick up more yield.

But IMO it is very clear that a good well run passive in terms of commodities fund can easily beat its benchmark index over time through prudent risk taking on term structure of the collateral and also of the futures curve . As I wrote in article recently both DFA and PIMCO have far outperformed their benchmarks and DFA is conservative on the collateral not venturing beyond 3 years and sticking with investment grade bonds. THE DJ UBS is designed to more evenly weight the individual commodities, so one can argue that it was designed with data mining if you like, but here I think it's hard to really make that case and at any rate its now live for 20 years.

while the DJ UBS has outperformed the GSCI by quite a bit, in portfolio either would have gotten same results, reason is that the higher SD is a POSITIVE when you have the type correlations we have here.

So bottom line is IMO I don't think your concerns are valid. Not much more I can say



Best wishes
Larry
Now as I said there are two good commodities indices which one can use to track the historical performance--one goes back to 70 and the other only to 1991 (DJ UBS). The GSCI is simple production weighted, so it's like a market cap, so cannot be any "data mining" to choose methodology. So IMO one can use that benchmark and will get a CONSERVATIVE answer as to how a well run CCF would have done because they have outperformed.
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

Larry,

This is fascinating. I think we actually agree about the observations, but perhaps disagree about the implications for investment, which is a minor point for me.
PCRIX is totally passive on commodities
You may be right about this, but that is not what PIMCO says in its description of the fund:
Combining the benefits of commodities with the experience of PIMCO as an active manager of commodities and fixed income collateral, the fund seeks to outperform the Dow Jones-UBS Commodity Total Return Index by actively managing both the commodities exposure and the underlying TIPS collateral portfolio.
http://investments.pimco.com/Products/pages/287.aspx
Looking at active funds tells you nothing about the performance of the passive strategy, so it's useless.
Not exactly. If one wanted to learn the collective returns to people invested in stocks, one could use the universe of ACTIVE managers to tell you that. One could not use a single active fund, but the collective returns are, by definition, the returns to active funds, minus of course what investors paid for the active "service".

Similarly, I argue that if one wanted to know collective returns to those invested in commodities, then the collective returns of the active managers (including, but certainly not limited to PCRIX) would tell you that. As with stocks, there would be a performance drag due to costs, and but there could not be a positive alpha from the COMMODITIES investing. Of course, there cannot be positive alpha to commodities investors overall, there have to be negative alpha losers as well. The commodities market certainly could have a positive Sharpe ratio, and a higher ratio than stocks or bonds.

If one were to look at the returns to COMMODITIES, as opposed to active management of an individual fund, then this would have to be cap weighted. By definition, since there is more money in oil than, for example, oranges. I can understand that an investor might wish to have her commodities exposure less dependent on energy than is the overall commodities market. However, that would mean that the investor, for whatever reason, wants to actively manage her exposure and NOT invest in the overall commodities market. But one could no more infer the returns to commodities investing by looking at the UBS index than one could infer returns to US stock investing by looking at the performance of an equal weighted index. Any one person, and any fund can invest in an equal weighted index. The market as a whole cannot invest equal weighted. The market as a whole is cap weighted, again, by definition.

The UBS index is neither cap nor equal weighted. It is formed annually according to rules designed to keep a variety of commodities representing a significant portion of the index. One can understand why they choose to do this without agreeing that this reflects returns to commodities. It tells us returns to this idiosyncratic mix of commodities. A mix that changes each year, not directly related to the composition of the commodities market. This also means that outperforming this particular index is not particularly interesting. For example, although PCRIX outperformed the UBS index, it underperformed the Morningstar Long Only Total Return index. But I would say "so what" to both observations. Both indices impose arbitrary weights on index composition and thus neither reflects the performance of the commodity market or the returns to commodity investors.

Since the returns to PCRIX are not simple returns to commodities investing, then they do not tell one what the returns to commodities investing would be. They include, to some unspecified extent, the returns to the underlying index, the returns to active trading to beat the index, returns to active management of commodities exposure to beat the index, and returns to the collateral investment strategy. Only the first reflects the returns to commodities. Whether the active returns will persist depends in part on chance , in part on how many other traders apply the same strategies, and in part on how the underlying index is composed. The collateral investment strategy is not a commodities return at all. It is real money, but it is not a return to commodities investing.

So... one can use the PCRIX as a hedge against stock and bond risk because one thinks that it is a good fund. One can use it because one believes its contribution to overall portfolio performance will remain positive- although the effect over the time you cited seems marginal at best. But one should recognize that the mean return came from active strategies and bond returns, not from commodities.

For me, this is fascinating, but the uncertainties about the performance of the commodities market as a whole, and uncertainties of the optimal passive strategy for investing in commodities leave me unprepared to dive in. It is not even clear that one can invest passively in commodities given the amount of trading required. At least, it may be impossible for an institutional investor to invest passively without being systematically exploited. An individual small investor perhaps could do so without attracting enough attention for others to front run him.

When there are data about returns to commodities investing, as opposed to the returns to one or two funds that do not track commodities, then at least one can have certainty about the size and sign of the commodities contribution. At the moment, I come back to "there are no data".

I suspect you agree with all of the above, but please correct me if I fail to understand your points.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
TO39
Posts: 203
Joined: Fri Mar 14, 2008 10:01 am

Re: How Commodities Can Help a Portfolio

Post by TO39 »

to afan, you said
As with stocks, there would be a performance drag due to costs, and but there could not be a positive alpha from the COMMODITIES investing. Of course, there cannot be positive alpha to commodities investors overall, there have to be negative alpha losers as well.

How does that explain this chart showing the inception of pcrix, a commodities index, vti, and inflation protected bonds.

http://quote.morningstar.com/fund/chart ... %2C0%22%7D

It looks like lots of alpha to me, for both pcrix and the index
afan
Posts: 8191
Joined: Sun Jul 25, 2010 4:01 pm

Re: How Commodities Can Help a Portfolio

Post by afan »

Perhaps getting too technical, but I was referring to commodities vs commodities, and the definition of alpha.

There cannot be positive alpha of commodities in aggregate to the commodities market. By definition. Just as a Total Stock index cannot have positive alpha to itself.

It would be possible for the commodities market to have higher returns than the stock market, and it would be possible for any given commodities fund to have higher performance than the stock market. However, these reflect different underlying assets, not alpha.

As I indicated, commodities as a whole, or an individual fund can have a higher Sharpe ratio, but this is not alpha, since they are invested in different things.

One could do the calculation of the returns of one asset vs another, and come up with a figure for "alpha" but it would be meaningless. During the crash, this would have shown a positive alpha for T bills, which would not reflect management contribution to a T bill portfolio, but just the fact that stocks went down and T bills did not.

In the example you cite, Morningstar is reporting the alpha of the fund against a commodity index. It is not reporting the alpha of commodities against commodities. There would be no point, since by definition, before expenses, the alpha has to be zero. You cannot compare the alpha reported for PCRIX to the alpha reported for VTI, since they are calculated against different indices.

Also, Morningstar uses the total return of the fund in calculating this alpha, not the component of the return that is due to its commodities exposure. So the performance of the fund does not tell you the performance of commodities investing, because it is not only invested in commodities, and it does not track the commodities market. See above.
We don't know how to beat the market on a risk-adjusted basis, and we don't know anyone that does know either | --Swedroe | We assume that markets are efficient, that prices are right | --Fama
Topic Author
larryswedroe
Posts: 16022
Joined: Thu Feb 22, 2007 7:28 am
Location: St Louis MO

Re: How Commodities Can Help a Portfolio

Post by larryswedroe »

afan
First, PIMCO is totally passive on the commodity side in terms of exposure to the commodities. Their active component is only trading on different exchanges to get better prices. So it's totally passive

Second, here is where Rick and I very much disagree. To him if it's not totally passive it's not good. That is bad thinking because there are negatives to pure indexing and thus well defined and well run funds can be created to beat over time indices as long as you are prepared to accept RANDOM tracking error, something Vanguard is not willing to accept to anything but a very minor degree. The active part is simply avoiding the key roll dates for the indices when the active traders know the indexers must trade and exploit them. Just doing that is probably worth 2% a year. Then using the cheapest to trade months is likely to add more, which is how DFA and PMCO have basically been able to outperform. Though in PMCOs case they also use TIPS and have had huge gains there as well and also more credit risk--which is why I stopped using them, plus DFAs lower fees.

Third, the returns to commodities IMO is easily known, in an investable sense by using the GSCI index--which is cap weighted, so there we disagree. And as I said if there had been well run funds they easily could have beaten the index so that the data is very conservative IMO as to the results investors could have earned.

Fourth, we totally disagree on the issue of active funds. While you are correct that it shows you what INVESTORS earned, that is irrelevant to whether one should invest in CCF or not because you would not choose an active strategy with its high costs and even loss of control over your exposures.

So the bottom line to me is simple that the GSCI is long data and fine to analyze--and gives you conservative estimates. It helps to decide if you want to include CCF or not. And the DJ UBS is also a good alternative with very logical and simple construction rules, which to me is what matters.

I hope that helps

Larry
Post Reply