Page 1 of 1

7 reasons to avoid commodities

Posted: Mon Jul 01, 2013 8:22 am
by larryswedroe
My review of the criticisms
http://www.cbsnews.com/8301-505123_162- ... yre-wrong/
Hope you find it helpful

Best wishes
Larry

Re: 7 reasons to avoid commodities

Posted: Mon Jul 01, 2013 9:32 am
by Browser
Reason #8: Even if you believe the commodities as diversifier story, the difference in portfolio returns and Sharpe between a portfolio with a small (5%-10%) allocation to commodities and one w/o commodities is small enough to be regarded as "noise" IMO. Why bother?

Re: 7 reasons to avoid commodities

Posted: Mon Jul 01, 2013 3:12 pm
by afan
Larry,

Thanks for summarizing many of the discussions you have lead here on Bogleheads. Most of your points seem well made. As discussed here, there is still the matter of how to evaluate a fund that seeks to track an index, but ends up with higher returns due to how it manages its collateral. As we have seen with commodities funds, the bursting of the bond bubble has hurt them. That means not much of a bond hedge, and it means that some of the out-performance was not due to better commodities futures performance.

I have asked this one before: If one wants to hedge stocks and bonds, why deal with the complexities of commodities, the unpredictable correlations, the trading costs, the swings into and out of contango, the taxes, etc.? Could you make an argument for straightforward hedging of stock and bond markets with futures or options on these markets? I don't invest in these derivatives, so excuse the rookie question. Why not just put a small amount in, for example, deep out of the money put warrants? Unlike commodities futures, you expect them to have negative long term returns, but you get very reliable hedging, perhaps at a lower cost. I have not tried to model how much one would have to invest, or how it would affect volatility of the portfolio, but it would be a pure hedge.

In the interest of full disclosure, I almost did this with Japanese stocks years ago. I refrained because "I don't invest in derivatives", but my wife has never forgiven me. We would have been on easy street.

On expenses, to be fair, the DFA can charge lower fees because the investor pays separately for some services that are included in the PCRIX expense ratio. A better comparison would require adding to the expense ratio for DFA some amount for the cost of the advisor required to get into funds. The 0.4% expense difference might be a good estimate of what going through an advisor might cost. Surely there are opportunities to get in for somewhat less, but many people pay much more.

Re: 7 reasons to avoid commodities

Posted: Mon Jul 01, 2013 3:18 pm
by Browser
afan makes some interesting points. Would like to hear more addressing them too, particularly the idea of using puts.

Re: 7 reasons to avoid commodities

Posted: Tue Jul 02, 2013 9:51 am
by Random Walker
Do TIPs in a portfolio partially offset the benefits of CCFs?

Dave

Re: 7 reasons to avoid commodities

Posted: Tue Jul 02, 2013 11:58 am
by pastafarian
I'm a skeptic at heart, and a non-believer in terms of the utility of CCFs in our portfolio. We don't own any. But that's not the point I'm aiming at. I'm just a retail DIY investor, so I offer no precise analytical rebuttal. Simply an observation, perhaps an unintended consequence of Larry's analysis.

I've owned Larry's books, and donated my copies to the local library. Over the years I experience an increasing sense (for lack of a better description) "investment cognitive dissonance" between his books, his forum posts and MoneyWatch blogs in terms of fixed income holdings. This article is one of those episodes. At the risk of being accused of being unfair to Larry's analysis, I'm focused on his fixed income component in Long Term Treasuries.

From Larry's Winning Bond Strategy I thought I learned that the slope of the interest curve is traditionally steepest at the short end; and that's where investors are rewarded. My take away (perhaps incorrectly) was investors should have a preference for short term high quality fixed income.

I seem to recall forum threads (several years ago) where Larry made an emphasis on TIPS dominating one's fixed income portion. That is except when real returns were below some threshold (with appropriate references to the St Louis Fed's TIPS charts). So I inferred perhaps something Larry did not imply. Investors should prefer TIPS except when real returns are low then they should prefer short term high quality fixed income.

I also think I recall a fairly recent thread where Larry emphasized long term treasuries in portfolios that are equity "heavy" (not defined). Which brings us to this blog in defense of CCFs. I suspect by his inclusion of LT Treasuries (in this blog), some DIY investors will infer Larry prefers them to ST Investment Grade or TIPS (and so should you!).

I know this does not address the utility of CCFs, I was simply observing what strikes me as an inconsistency between Larry's FI advice over the years. I appreciate that Larry is a FI expert and executes trading strategies to optimize his clients portfolios.

I reckon there's a caveat for everything. What's a DIYer to make of it? :beer

Re: 7 reasons to avoid commodities

Posted: Tue Jul 02, 2013 1:32 pm
by nisiprius
Larry, I'm not sure that I see where or how your column addresses the (harsh) criticism that William J. Bernstein presents in Skating Where The Puck Was. I hate to try to summarize, but I will try anyway. Bernstein says something like this, but I may be distorting or caricaturing. He thinks entry of new investors into the commodity futures market, with different goals from the old investors, has changed the characteristics of the market. "Rekenthaler's rule: if [retail investors like Nisiprius] know about it, it doesn't work any more." In detail, he says:
  • Performance before 1990, "when Gibson wrote the first edition of Asset Allocation," shouldn't really count because commodities were all but non-investible. "Their major players were folks who were more concerned with simple survival than in collecting an attractive Markowitz grid."
  • Now "the biggest players are the likes of PIMCO and Oppenheimer, who have sold the gullible on the pre-1990 record of commodities futures."
  • The characteristics of the market have changed, because the purpose of commodities futures investors is now "supposedly, protection against inflation, not deflation as had been the case prior to 1990."
  • Therefore, the process has shifted from "normal Keynesian backwardation... [to contango]."
  • Because of the change in the investor population and their reasons for investing, "during the 2007-2009 financial crisis, the other shoe dropped; commodities futures, rather than being a non-correlating 'alternative' asset, exhibited price declines just as devastating as those in equities."
He exhibits these charts in evidence. The first shows the falling roll return, the second shows the increasing correlation.

Image

Image

Re: 7 reasons to avoid commodities

Posted: Tue Jul 02, 2013 6:57 pm
by Browser
Wonderful post, Nisiprius. These are really some major issues re: CCFs that investors need to resolve in their own minds before swimming in. Frankly, I'm persuaded by Bernstein's arguments because they're forward-looking, and rather skeptical that the backtesting data supporting the CCF case is enough to hang one's hat on. You have to be able to stick with whichever choice you make over the long run, or it doesn't count. The same arguments can be applied to investing in gold, since now the Rekanthaler investors have ready access via the numerous bullion ETFs out there.

Re: 7 reasons to avoid commodities

Posted: Tue Jul 02, 2013 7:23 pm
by Bradley
nisiprius wrote:........ I hate to try to summarize, but I will try anyway. Bernstein says something like this, but I may be distorting or caricaturing. He thinks entry of new investors into the commodity futures market, with different goals from the old investors, has changed the characteristics of the market.
  • Performance before 1990, "when Gibson wrote the first edition of Asset Allocation," shouldn't really count because commodities were all but non-investible. "Their major players were folks who were more concerned with simple survival than in collecting an attractive Markowitz grid."
  • Now "the biggest players are the likes of PIMCO and Oppenheimer, who have sold the gullible on the pre-1990 record of commodities futures."


    He exhibits these charts in evidence. The first shows the falling roll return, the second shows the increasing correlation.

    Image

    Image

nisiprius,

Bill has been consistent over the years in his concern over the data used by academics to support the use of CCFs. Conclusions drawn from passive indices whose allocation have varied over the years vs actual live managed CCFs with real world issues including expenses deserve little attention IMO. The charts you posted also reveal the short comings of much of the academic works supporting CCFs. Academics,relying on relatively short time periods of live funds to determine long term correlations and return data may produce inaccurate data/conclusions.

Bradley

Re: 7 reasons to avoid commodities

Posted: Sun Jul 07, 2013 9:23 pm
by Ranger
nisiprius wrote:
Image
I am not able to reproduce the chart above from the data. Here is the correlation data chart from the source data.

Image

SPX and TNX data is from yahoo finance, DJUBS total return index data from Dow Jones website.
http://www.djindexes.com/mdsidx/downloa ... l_hist.xls"

My chart is some what consistent with other charting package charts and gives completely different picture than what Bernestein is advocating

EDIT: I am assuming when he refers to commodities, he is using DJUBS index. He could be using other indices like CRB. Also i am using ten year notes instead of total bond market index.

Re: 7 reasons to avoid commodities

Posted: Mon Jul 08, 2013 1:30 pm
by Bradley
The following quote from Skating Where The Puck Was by William J Bernstein may be helpful.

“ ...........During the 2007-2009 financial crisis, the other shoe, in the form of much higher correlations dropped; commodities futures, rather than being non-correlating “alternative” asset, saw prices decline just as devastating as those in equities. Figure 5 shows how during the 2007-2009 crisis, the correlation between stocks, as represented by the S&P 500, and commodities futures had increased dramatically and has remained high since. The only saving grace is that the correlation between bonds, represented by the Barclays/Lehman Aggregate Index, and commodities has remained low. At the present time, then, the best that can be said about commodities futures is that they offset the risk of longer bonds, if that’s an asset class that appeals to you.”

This booklet shows how the world is becoming much more connected with resultant higher global correlations.

Re: 7 reasons to avoid commodities

Posted: Mon Jul 08, 2013 3:33 pm
by staythecourse
Bradley wrote:The following quote from Skating Where The Puck Was by William J Bernstein may be helpful.

“ ...........During the 2007-2009 financial crisis, the other shoe, in the form of much higher correlations dropped; commodities futures, rather than being non-correlating “alternative” asset, saw prices decline just as devastating as those in equities. Figure 5 shows how during the 2007-2009 crisis, the correlation between stocks, as represented by the S&P 500, and commodities futures had increased dramatically and has remained high since. The only saving grace is that the correlation between bonds, represented by the Barclays/Lehman Aggregate Index, and commodities has remained low. At the present time, then, the best that can be said about commodities futures is that they offset the risk of longer bonds, if that’s an asset class that appeals to you.”

This booklet shows how the world is becoming much more connected with resultant higher global correlations.
I respect Dr. Bernstein depth of knowledge, but anyone expecting commodities to do well in a financial recession just doesn't get it. Think back to 2008 and ask yourself when the economy was slowing, folks were getting fired, Bears Stearns and others were going down in flames did ANYONE sit there and go, "no problem my futures contract in pork bellies or cotton will save me." No of course not. That is the only way commodities would have helped at that time.

It
was and is obvious commodities would not have helped then. If the economy slows down WORLDWIDE how in the world how would commodities do well??

This is another situation of what I always say is the faults of ALL investors. When folks come up with their portfolio it is their DUTY to understand the adv. and disadv. of EVERY asset class and only then decide if it has a role in your portfolio.

Here is when commodities will do well: 1. Real or possible supply- demand shock. 2. Hedge in the immediate period of UNEXPECTED inflation.

Any other time futures will still help with its low correlation to paper assets (stocks and bond) albeit with lower returns. So no surprise adding commodities without the two points above occurring (last 20 yrs.) gives a portfolio of stocks and bonds a decrease in volatility secondary to noncorrelation at the expense of lower return.

As I have said before the real use of adding commodities should be with a bond heavy portfolio where its real impact would ideally help.

Good luck.

Re: 7 reasons to avoid commodities

Posted: Mon Jul 08, 2013 4:16 pm
by Ranger
Bradley wrote:The following quote from Skating Where The Puck Was by William J Bernstein may be helpful.
I am not sure, whether you are replying to my post or just posting some random quotes from the book.

I was just pointing out Dr. Bill Bernestein's data is wrong and hence conclusion that was drawn from it. According to him, as pointed out by nisiprius, markets have changed because of all these new investors
"
nisiprius wrote:Larry, I'm not sure that I see where or how your column addresses the (harsh) criticism that William J. Bernstein presents in Skating Where The Puck Was. I hate to try to summarize, but I will try anyway. Bernstein says something like this, but I may be distorting or caricaturing. He thinks entry of new investors into the commodity futures market, with different goals from the old investors, has changed the characteristics of the market. "Rekenthaler's rule: if [retail investors like Nisiprius] know about it, it doesn't work any more." In detail, he says:
.....
[*]Because of the change in the investor population and their reasons for investing, "during the 2007-2009 financial crisis, the other shoe dropped; commodities futures, rather than being a non-correlating 'alternative' asset, exhibited price declines just as devastating as those in equities."[/list]

He exhibits these charts in evidence. The first shows the falling roll return, the second shows the increasing correlation.
From my chart, it clearly shows markets have not changed as it still showing same behavior as before. If you are referencing his quote stating correlations of all risky assets go to one, during the time of stress. then sure, it is just common sense.

Re: 7 reasons to avoid commodities

Posted: Mon Jul 08, 2013 5:49 pm
by Bradley
Ranger wrote:
I am not sure, whether you are replying to my post or just posting some random quotes from the book.

I was just pointing out Dr. Bill Bernestein's data is wrong
Ranger,

Just finished reading Bernstein's booklet and thought it might be helpful to add some context/insight into his thinking by posting a snippet that expanded on earlier original post from booklet and was not meant to be a response to your post. What specifically is wrong with Bill's data?

Re: 7 reasons to avoid commodities

Posted: Mon Jul 08, 2013 9:43 pm
by Ranger
Bradley wrote: Ranger,

Just finished reading Bernstein's booklet and thought it might be helpful to add some context/insight into his thinking by posting a snippet that expanded on earlier original post from booklet and was not meant to be a response to your post. What specifically is wrong with Bill's data?
I haven't read the book, so I can't comment on other than the "Figure 5" posted on this thread. Figure 4 mentions DJUBS, so I am assuming he is using DJUBS data for Figure 5 also. I tried to recreate the same graph using data from direct source (Dow Jones) and Stock data from (SPX) from yahoo finance. I have posted my graph, which should be similar to Figure 5, but it is not. I have verified my graph from couple of other places.

In my opinion, he is wrong with the conclusion that markets have changed considerably and shows increased correlation because of "new retail investors" is based on graph that is wrong and cannot be recreated. If you reference my chart, increased correlation didn't just start in 2007-09, it was there in 1995, 2000, 2003 even before "new retail investors" piled on to these CCF's.

Further more, if those retail investors still there in CCF's, why correlation has come back as of today?

Re: 7 reasons to avoid commodities

Posted: Tue Jul 09, 2013 11:56 am
by Bradley
Ranger wrote:I tried to recreate the same graph using data from direct source (Dow Jones) and Stock data from (SPX) from yahoo finance. I have posted my graph, which should be similar to Figure 5, but it is not. I have verified my graph from couple of other places.

In my opinion, he is wrong with the conclusion that markets have changed considerably and shows increased correlation because of "new retail investors" is based on graph that is wrong and cannot be recreated.


Ranger

If you have access to Rick Ferri’s All About Asset Allocation, compare the 36-Month(3yr) correlation Figure 10-6 on page 181 to Bernstein’s Figure 5 Rolling 3-Year correlation figure. Their data points are very similar, ( even though Rick uses CRB data and Bill appears to use DJ_UBS data ) which IMO raises the confidence level in their charts/data being accurate. Your chart appears much more volatile than other rolling correlation charts. Maybe someone experienced with rolling correlation charting can help explain the disparity between your chart and the published charts referred to above.

Bradley

Re: 7 reasons to avoid commodities

Posted: Tue Jul 09, 2013 2:05 pm
by Ranger
Bradley wrote: Ranger

If you have access to Rick Ferri’s All About Asset Allocation, compare the 36-Month(3yr) correlation Figure 10-6 to Bernstein’s Figure 5 Rolling 3-Year correlation figure. Their data points are very similar,
Bradley
This will be my last post in this thread, since this is already tangential to the original OP topic.

No, I do not have Rick's book either. I have provided with you the source data of DJUBSTR in my earlier posts. It is pretty to easy to calculate rolling correlation in excel or Google Spreadsheet. SPX is data is available freely from yahoo or other websites.

Here is the source data for $CRB.

http://www.jefferies.com/Commodities/2cc/389

Here is another easy chart from stockcharts. (click to enlarge)

Image

Here is one for $DJUBSTR. Click to enlarge

Image

They are not that different from my excel charts.

Finally i am not an academic. I am an low life short term trader. These 36 months correlation are useless to me. For me 36 minute correlation is far more important than these graphs. Hopefully if i am wrong, some body will let me know.

Thx

Re: 7 reasons to avoid commodities

Posted: Tue Jul 09, 2013 9:38 pm
by Ranger
Even tho' I said previous post was my last post, I just wanted to address central criticism of wonderful Larry's article in this thread is echoed from Dr. Bernestein's book.
nisiprius wrote: He thinks entry of new investors into the commodity futures market, with different goals from the old investors, has changed the characteristics of the market. "Rekenthaler's rule: if [retail investors like Nisiprius] know about it, it doesn't work any more." In detail, he says:
  • Performance before 1990, "when Gibson wrote the first edition of Asset Allocation," shouldn't really count because commodities were all but non-investible. "Their major players were folks who were more concerned with simple survival than in collecting an attractive Markowitz grid."
  • Now "the biggest players are the likes of PIMCO and Oppenheimer, who have sold the gullible on the pre-1990 record of commodities futures."
  • The characteristics of the market have changed, because the purpose of commodities futures investors is now "supposedly, protection against inflation, not deflation as had been the case prior to 1990."
  • Therefore, the process has shifted from "normal Keynesian backwardation... [to contango]."
  • Because of the change in the investor population and their reasons for investing, "during the 2007-2009 financial crisis, the other shoe dropped; commodities futures, rather than being a non-correlating 'alternative' asset, exhibited price declines just as devastating as those in equities."
He exhibits these charts in evidence. The first shows the falling roll return, the second shows the increasing correlation.
These are the same general criticism found in some mainstream media and investors.

Dr. Bernestein has been clearly wrong. Contago didn't persist even tho there are plenty more investors and short term traders in the CCF area. Take a look at future curves of one of the most actively traded commodity market, which is oil. I have also included Gasoline and Heating Oil.
(click to enlarge)
Image
Image
Image


Number of contracts traded and open interest in these futures are at record level, but these contracts are in backwardation till 2020. This is more repudiation of his theory than correlation drift, which i presented earlier.