Diversification Benefits of PCRDX Tested

[Gone4Good]

Hi all,
Over the past several days (weeks?) I've been in numerous discussions about the merits of diversifying ones portfolio with some commodity exposure.  If you're all getting tired of it, fear not, so am I.  Not withstanding my fatigue, I spent some time doing some scenario analysis using the portfolio tools at the following site:

www.riskgrades.com

If you haven't been to RiskGrades before, its worth a gander.  The free site allows you to test the diversification benefits of a portfolio and run some scenario analysis to see how changes in the portfolio affect its overall risk. 

Considering how new the PCRDX fund is, I'm not sure how valid the results are.  But here is what I found when adding about 10% PCRDX exposure to my existing portfolio:

1)  PCRDX alone carries a "risk grade" of 111 which means that only "1% of 'US Funds' are riskier than PCRDX" (please see the site for definitions).  The Vanguard S&P 500 Index fund carries a "risk grade" of 41, which means that "46% of 'US Funds' are riskier than VFIAX"

2)  When adding PCRDX by selling equity funds, my portfolio risk grade went down (meaning the portfolio is LESS risky).

3)  When adding PCRDX by selling bonds, my portfolio risk grade went UP (meaning the portfolio is MORE risky).

4)  When adding PCRDX by selling money market funds, my portfolio risk grade went UP.

The conclusion, at least according to RiskGrades, is that PCRDX can reduce the risk of a portfolio if it replaces equity exposure but will increase the risk profile of the portfolio if it replaces fixed income investments.

Cheers.

 

[Dry Socks]

Yrs, I read some of the prior posts...

These results sound very reasonable if PCRDX is highly volatile and negatively correlated with the stock market. I think that's the case.

Thanks for performing the research !

-Dave

 

[wildcat]

Good stuff Yrs.  It works for me b/c I am young & heavy on the equity side.

Vanguard has a nice article about inflation hedges and it covers commodities.

 

[saluki9]

Good stuff Yrs. It works for me b/c I am young & heavy on the equity side.

Vanguard has a nice article about inflation hedges and it covers commodities.

I've read a lot of research on this subject. I tend to go both ways depending on what day it is. Most of the research shows that commodities in general benefit people with a more conservative portfolio, but become less valuable as you increase your equity exposure. (this is what they taught in the CFA materials too)

Then I've read some stuff from DFA and seen a presentation by Ken French in which he proves (rather convincingly that they are worthless and you're better off taking your risks with equity)

I still hold about 5% of my portfolio in PCRDX

 

[wildcat]

The Vanguard article basically said commodities can be moderately effective in terms of inflation especially if based on futures contracts & if some sort of collateral is used like TIPS.  The best part of commodities exposure is the low to negative correlation with other investments.  However, the article really questions whether or not it is worth the volatility.

Stocks (as we know) are still king when it comes to beating inflation but again more risk & volatility.

But as Saluki9 pointed out I have read both sides of the story & I am comfortable with a small position.  I don't expect much other than a good hedge when everything is crappy. 

 

[Gone4Good]

Some more tinkering with RiskGrades and some hypothetical portfolios comprised of 70% Vanguard Total Stock Market (VTSMX) and up to 30% in Vanguard Bond Index (VBMFX), Vanguard REITs (VGSIX) or PIMCO Real Return Commodity (PCRDX) yielded the following results:

(Lower risk grades reflect lower volatility) 

Portfolio                                                                                Risk Grade
VTSMX 70%, VGSIX 30%                                                         53
VTSMX 100%                                                                         50
VTSMX 70%, PCRDX 30%                                                       48
VTSMX 70%, PCRDX 15%, VGSIX 15%                                        47
VTSMX 70%, VBMFX 15%, VGSIX 15%                                       44
VTSXM 70%, PCRDX 10%, VGSIX 10%, VBMFX 10%                      42
VTSMX 70%, VBMFX 15%, PCRDX 15%                                       39
VTSMX 70%, VBMFX 30%                                                       35

So it would appear that with respect to reducing portfolio volatility - bonds still rein supreme (at least according to risk grades). Naturally, the expected return for each of the above portfolios is different, so it is still not clear how much potential upside is sacrificed for the benefit of reduced volatility. I still question the long-run returns generated by lumps of inanimate objects. So while PCRDX may reduce the volatility of heavily equity weighted portfolios, a portfolio with an equal mix of bonds results in even lower volatility and may also produce higher expected returns.

ronin, I have to admit I'm not familiar with sortino's work, but it would seem to me that we're all talking about similar things. If I construct two portfolios with similar "expected returns", both of which meet my return objectives, the lower volatility portfolio seems to be a clearly better choice. I would also think that the lower vol portfolio is the one with the lowest risk of underperforming my "MAR" or "expected return". No?

 

[saluki9]

The point Ken French was making during his presentation is that commodities futures are a zero sum game unlike stocks. Therefor the total return of all holders is 0% minus transaction costs which will make it negative.

 

[brewer12345]

The point Ken French was making during his presentation is that commodities futures are a zero sum game unlike stocks.  Therefor the total return of all holders is 0% minus transaction costs which will make it negative.

That's all fine and well, except that many of the players in the futures market are not there to make money, but to lose it. Commodity producers and consumers are not in the futures markets to make money. They are there to lay off commodity risk. When you lay off risk, you virtually always pay up. Guess who is taking on the risk in return for the reward? That's right, investors/speculators. The fact that it is a zero sum game doesn't really matter to us as investors.

 

[saluki9]

That's all fine and well, except that many of the players in the futures market are not there to make money, but to lose it. Commodity producers and consumers are not in the futures markets to make money. They are there to lay off commodity risk. When you lay off risk, you virtually always pay up. Guess who is taking on the risk in return for the reward? That's right, investors/speculators. The fact that it is a zero sum game doesn't really matter to us as investors.

French and Fama both completely refuted that point. I don't want to threadjack, but if you're interested I can go dig up my materials from the presentation

 

[brewer12345]

French and Fama both completely refuted that point.  I don't want to threadjack, but if you're interested I can go dig up my materials from the presentation

Eh, I'm not going to buy it, so don't bother unless it is pretty easy.

 

[saluki9]

Eh, I'm not going to buy it, so don't bother unless it is pretty easy.

I found it. The argument from Fama and French is that the short supply argument applied in the 1950's but doesn't work any more. Most of the contracts traded and held by the funds are notional in nature and no longer backed by physicals.

The other argument was that the data is MUCH too short. The only reputable data is available from 1959 and the market wasn't very liquid at that point.


Ken French and Eugene Fama are pretty smart guys, so I give them the edge on that (BTW: I think they got hosed again today by not winning the nobel prize in econ for which I think they are long overdue)

The whole presentation is 30+ pages, but the present many good arguments against investing in commodities.

 

[Gone4Good]

The point Ken French was making during his presentation is that commodities futures are a zero sum game unlike stocks.  Therefor the total return of all holders is 0% minus transaction costs which will make it negative.

In an efficient market this is true for "unfunded" futures trades.  PCRDX is a "funded" or "collateralized" investment vehicle, which should give investors the same return as the underlying basket of commodities assuming markets are efficient.  That is not to say those returns are "good" they're just not necessarily zero.  

(see here for some commodity graphs) http://early-retirement.org/forums/index.php?topic=4404.msg76545#msg76545

Most of which are pretty close to zero though  

That's all fine and well, except that many of the players in the futures market are not there to make money, but to lose it.  Commodity producers and consumers are not in the futures markets to make money.  They are there to lay off commodity risk.  When you lay off risk, you virtually always pay up.  Guess who is taking on the risk in return for the reward?  That's right, investors/speculators.  The fact that it is a zero sum game doesn't really matter to us as investors.

brewer - I disagree with both you and Greer on this.  Futures markets are not insurance markets.  Contrary to the contrived cattle markets example in Greer's research BOTH producers AND consumers have an interest in laying off their commodity price exposure.  Do you really think jewelers don't hedge their gold exposure?  And if they didn't, why wouldn't speculators buy the spot commodity and sell the forward strip to extract that "insurance premium" at no risk?

Even if there is an excess number of natural forward sellers relative to the number of buyers how do you think the new (long biased) commodity funds like PCRDX will impact that balance?  It's possible that as these funds grow they could tip the balance the other way such that YOU are paying the insurance premium

 

[brewer12345]

brewer - I disagree with both you and Greer on this.  Futures markets are not insurance markets.  Contrary to the contrived cattle markets example in Greer's research BOTH producers AND consumers have an interest in laying off their commodity price exposure.  Do you really think jewelers don't hedge their gold exposure?  And if they didn't, why wouldn't speculators buy the spot commodity and sell the forward strip to extract that "insurance premium" at no risk?

Even if there is an excess number of natural forward sellers relative to the number of buyers how do you think the new (long biased) commodity funds like PCRDX will impact that balance?  It's possible that as these funds grow they could tip the balance the other way such that YOU are paying the insurance premium

I'm pretty sure I never said that futures markets were insurance markets.  I'm pretty sure that
I indicated my belief that they have elements of insurance.

Speculators don't typically buy on the spot market because they don't want the physical commodity.  You really think some commodity pool operator wants a delivery of 100,000 bushels of corn?

Yes, of course both consumers and producers hedge their exposure in the futures markets.  However, there will always be a mismatch (one way or the other), and where there are mismatches there is return to be doled out to capital willing to take the risk on.  No doubt there will eventually be a situation where the futures market becomes popular enough that the risk premium gets squeezed out of being long commodity futures.  When it happens, it will be worth re-assessing the allocation.  I think we saw a brief instance of exactly that within the past year when the oil market went from forwardation to backwardation.  If I start seeing commodities markets consistently backwardated, I will be cashing out.

I have to say that I haven't seen anything presented here that refutes the academic research I have read and raddr's more pragmatic work.  It appears that an unlevered commodities portfolio has negative correlation with equities and FI and has significant positive return and volatility.

 

[Gone4Good]

However, there will always be a mismatch (one way or the other), and where there are mismatches there is return to be doled out to capital willing to take the risk on.  

But PCRDX isn't a hedge fund looking to extract "excess returns" on both sides of the market. It is a long biased (only?) fund. If you believe there are excess returns because of an imbalance between buyers (shortage) and sellers (excess) in the futures market you also have to accept the possibility that your fund may one day find itself on the other side of that equation.

It appears that an unlevered commodities portfolio has negative correlation with equities and FI and has significant positive return and volatility.

Significantly positive return? I haven't seen ANY evidence of that.

 

[wildcat]

How bout we call it a draw and to each his/her own?  Interesting discussion though.

One question on excess returns. Couldn't it be a result of one side having superior information?

 

[brewer12345]

Significantly positive return?  I haven't seen ANY evidence of that.

Go do some reading:

http://raddr-pages.com/RetireEarlyWeb/CommodityFutures.htm

I would guess that you have already made uyp your mind, though, so I am happy to agree to disagree, as wildcat suggests.

 

[ESRBob]

For those who do come down on the side of owning Commodities, wouldn't you rather be in PCRIX than PCRDX? (I thought they were the same except that PCRIX has a fee about a half percent lower per year). Didn't we work out last year that some of the big fund family supermarkets would let you buy into PCRIX at small (2.5k or so) minimum purchases?

 

[wildcat]

I think some don't have access to it cheaply via Vanguard for example (which I believe has since raised minimums on PCRIX).

 

[Gone4Good]

Go do some reading:

You're funny!

The chart looks nice, and it is better than I would have thought.  But the average annualized return from 1982-2005 is about 4.7% - not too bad but hardly what I would call "significantly positive."  If you lay the S&P 500 on the same chart over the same time period it would end at about 1,000 vs. 300.  Given where interest rates were over that time period you could have done as well or better in a money market fund.  I also think the returns shown here for RJ/CRB are better than those of DJAIG. 

I will agree to disagree with you both on this one. 

just for some context:

 

[haha]

You're funny!

The chart looks nice, and it is better than I would have thought.  But the average annualized return from 1982-2005 is about 4.7% - not too bad but hardly what I would call "significantly positive."  If you lay the S&P 500 on the same chart over the same time period it would end at about 1,000 vs. 300.  Given where interest rates were over that time period you could have done as well or better in a money market fund.  I also think the returns shown here for RJ/CRB are better than those of DJAIG. 

I think these comparisons are very sensitive to the period chosen. For example, the ten years prior to 1982 would have told a different tale.

Ha

 

[Spanky]

Yes, a different period would have shown a totally different result. The point is that their correlation is low.

 

[brewer12345]

For those who do come down on the side of owning Commodities, wouldn't you rather be in PCRIX than PCRDX?  (I thought they were the same except that PCRIX has a fee about a half percent lower per year).  Didn't we work out last year that some of the big fund family supermarkets would let you buy into PCRIX at small (2.5k or so) minimum purchases?

I'd love to have access to PCRIX, but Schwab doesn't offer it to cheesy retail accounts and I can't justify opening up another account with someone else just t get access to pcrix.

 

[ESRBob]

Yeah,
Either they changed the minimum at Vanguard or I am just mistaken -- the minimum there now for both IRA and regular brokerage accounts through Fund Access is 25k, not the 2.5k I thought I remembered. Sorry.

 

[cute fuzzy bunny]

Am I mistaken or doesnt PCRIX/RDX use TIPS as collateral for the commodities contracts purchased, therefore giving one the returns from TIPS as well as the commodities returns?

If thats the case, then the chart above only shows slightly less than half of the returns. Add an average of 3% a year for historic CPI and a couple of percentage points for the tips coupon.

My further understanding is that PCRIX and commodities in general give you superior protection from UNEXPECTED inflation. I'm not entirely sure what that means in detail, but it sure makes ya feel all warm and fuzzy.

After having seen plenty of very smart people over at Raddrs site do very detailed analyses, as pointed to by Brewer, it appears that adding a slug of commodities via PCRIX reduces volatility while increasing returns and SWR's. Adding a lot appears to statistically do you even better. I'm talking out of my pants pocket, but IIRC a portfolio of about half small cap value and half PCRIX produces a SWR in the 6-7% range, backtested, admitedly not very far back.

I dont have the big brass ones to go that far. Still pondering some PCRIX although every time I look at it, it seems to be too expensive and to have had a big runup, but when I look at it a few months later, its up even more. Go figure.

I'm not sure what Fama and French had to say in detail about commodities, but I doubt they were talking about an instrument like PCRIX/PCRDX. Thats sort of like taking someones opinion on NASDAQ stocks since 2000 and correlating that to ones overall opinion of equities over the last 100 years.

 

[wildcat]

therefore giving one the returns from TIPS as well as the commodities returns?

Yup TIPS + Net gain on futures contracts

I didn't time (I did luck out though) b/c I wanted to be in the long haul