Home > Bond Market, Commodities, Investing, Trading > It’s Risk Parity, Not Risk Party

It’s Risk Parity, Not Risk Party

I am disinclined to take victory laps when most people are losing money, but the recovery in commodities prices over the last week at the same time that bond and equity prices are both declining is a taste of success for my view that has been rare enough lately. That is, of course, the burden that a contrarian investor bears: to be wrong when everyone else is having fun, and to be right when no one wants to go out and celebrate. In fact, if you find yourself sharing your successes too often with other people who are having the same successes, I would submit you should be wary.

It is worth noting that the commodities rally has not been led by energy, despite the terrible violence in Egypt which threatens, again, to ignite a spark in the region. Today, the rise in commodities was led by gold and grains; yesterday by cows and copper (well, livestock and industrials).

I don’t think that this is because of a sudden epiphany about inflation. In fact, although breakevens have been recovering from the oversold condition in June (more on that in a moment), the inflation data today did nothing to persuade inflation investors that more protection is needed. I gave some thoughts about the CPI report earlier today in this post, but suffice it to say that it was not an upside surprise. (And yet, there are starting to appear more-frequent smart articles on inflation risks. I commend this article by Allan Meltzer to you as being unusually clear-eyed.)

And commodities are not moving higher because of renewed enthusiasm about growth, I don’t think. Today economic bell cow Wal-Mart cut its profit forecast because higher taxes are causing shoppers to be more conservative (perhaps in more ways that one). And, while today’s Initial Claims figure was good news (320k versus expectations for 335k), weakness was seen in Industrial Production (flat, with downward revisions, versus expectations for +0.3%) and both Empire Manufacturing and Philly Fed came in slightly weaker than expectations. None of this is apocalyptic, but neither is it cause for elation about domestic or global growth prospects.

While the nascent commodity rally makes me personally feel warm and fuzzy, the more-momentous move is in what is happening to interest rates. And here I need to recognize that until very recently, I thought that bonds would follow the typical pattern of a convexity-exacerbated selloff: after a rapid decline, the market would consolidate for a few weeks and then recover once the overhang had cleared. I’ve seen it aplenty in the past, and that was the model I was operating on.

But I believe rates are heading higher. Although the overhang from the prior convexity selloff has probably been distributed, there is a new problem as illustrated by the news today about Bridgewater’s “All Weather” fund. The All-Weather Fund is an example of a “risk parity” strategy in which, in simplified form, “low-volatility” strategies are levered up to have the same natural volatility as “high volatility” strategies. The problem is that levering up an asset class with a poor risk-adjusted return, as fixed-income is now, doesn’t improve returns or risks of the portfolio at large. The -8% return of the AWF in Q2 illustrates that point, and makes clear to anyone who bought the great marketing of “risk parity” strategies that they probably have much more rate risk than they want (although according to the Bloomberg article linked to above, Bridgewater “hadn’t fully grasped the interest-rate sensitivity” of being long 70% of net assets in inflation-linked bonds and another 48% in nominal bonds. I do hope that’s a mis-quote).

The unwinding of some of that rate risk (Bloomberg called the panicky dumping of a relatively cheap asset class, TIPS, into the teeth of a retail and convexity-led selloff “patching” the risk) helped TIPS bellyflop in May and June, and to the extent that institutional investors wake up and reduce their levered long bets on fixed income we might see lower prices much sooner than I expected across the entire spectrum of fixed-income. Indeed, without the Fed or highly levered buyers, it’s not entirely clear what the fair clearing price might be for the Treasury’s debt. I was at one time optimistic that we would get a bounce to lower yields after a period of consolidation, but this news is potentially a game-changer. Although the seasonal patterns favor buying bonds in August and early September, the potential downside is much worse than the potential upside.

  1. Dan
    August 21, 2013 at 12:02 am


    I’ve got a question unrelated to this piece (sorry), but I enjoy your non-wild-eyed commentary, so I thought I’d ask: What’s your view of using GDP deflator vs CPI when it comes to calculating real GDP growth? I just saw this piece by Peter Schiff: http://www.europac.net/commentaries/gdp_distractor

    Personally, I think the inflation figures that come out of ShadowStats is a big steaming pile of nonsense, but I think the point that Schiff raises is worthwhile. So, why isn’t CPI used when calculating real GDP growth?


    • August 21, 2013 at 5:52 pm

      Well, for starters he’s not quite right (as Schiff, when he is writing about inflation, often is…in the ballpark, but not quite right). GDP is broader than CPI, because it includes (among other things) the government sector (which is not “consumer,” ergo not covered by the CPI). GDP has no more “flexibility to swap weights” than the CPI does; it just has different weights by virtue of the fact that is covers EVERYTHING, and so as the composition of total GDP changes, so do the weights in the deflator. By definition it must be this way, because Amount of Stuff * Price of those things = total dollar amount produced, which is GDP.

      So if you took the total dollar amount of stuff produced, and divided by a price index associated with consumption, you would have a badly biased growth number since you would not be considering how the inflation rate is different for Investment, Government, and Net Exports (the parts, along with Consumption, in GDP).

      Now, the GDP deflator is probably a better number for measuring the aggregate price level than is the CPI. But as consumers, we tend to care more about the price level of stuff we actually buy directly, which is what CPI measures.

      The reasons that GDP and CPI diverge are similar to the reasons that PCE and CPI diverge, except more so – too much medical, for example, and not enough housing, compared to what the average person consumes; or you could say that CPI underweights medical and overweights housing compared to the total production of these things in the GDP. Take your pick, neither position is incorrect! (However, you CAN make an error by using one of them for the wrong purpose. They are not interchangeable – you have to know why you’re using the one you’re using!)

      If the government sector was much smaller, the differences would be much lower. In particular, if the government was a smaller player in healthcare, the differences would be smaller. But as the government grows as a % of GDP, so does the difference in the baskets between these two measures.

      Hope that helps!

      • Dan
        August 24, 2013 at 1:52 pm


        Thanks for your explanation; it does make sense. However, it has me wondering how public sector inflation is measured for PCE… After all, it certainly doesn’t feel like government is getting any cheaper! If anything, sectors with high government involvement — I’m thinking about education and medicine — tend to have an awful lot of inflation. (But I suppose taxes are adequately covered by CPI?) It’s probably difficult to measure the cost of government services.

        Although I think we should always be concerned about any consumer price inflation, I feel that both the conspiracy theorists and the “there’s no inflation here” folks miss the point. As someone who needs to put his family’s money to work to make sure we all have roofs over our heads in future, I see massive inflation in the asset markets. Maybe the price of a Big Mac isn’t rising so fast, but it certainly costs a heck of a lot to buy a series of future cash flows! As I tried to explain to a relative who’s comfortably retired from the civil service, “I don’t celebrate when Safeway raises the price of steak, so I why should be happy when the stock market rises? It just means that dividends cost more.” His reply was a clueless blank face, but it wouldn’t matter anyway — he’s a defined-benefits pensioner backstopped by the taxpayer!

        Thanks again,

      • August 24, 2013 at 3:48 pm

        That’s exactly right…most people should dislike rising asset markets, unless they are closer to the distribution phase of their lives. But young accumulators often seem to LOVE rising markets. It’s absurd!

        FYI, taxes are not a consumption item, so are not included in either PCE or CPI.


  2. HP Bunker
    August 29, 2013 at 12:11 pm

    Hi Mike, Don’t know if you saw the following:


    I mention it not for the focus on Section 8 renters, but rather for some intriguing stats showing the disparity between home price increases and rents (for example, +19% vs. +2% over the past year in the Atlanta area). Like I’ve been saying, houses behave just like three-dimensional bonds these days (ie price up –> yield down)! If tapering really happens in the next few months, the divergence between home prices and rents could narrow somewhat. Rising mortgage rates should have that effect also. But the trend over the past year or so, while strange from a historical perspective, makes sense in light of the extremely cheap money that’s been available and generally dismal investment options.

    • August 29, 2013 at 1:34 pm

      I didn’t see that, since I have been on vacation – so thanks! But I will point out that there is nothing unusual about the disparity between home price increases and rents. Rents always lag, by roughly the lag that currently exists. It will be unusual if six months from now we haven’t seen rents rise, but it’s not unusual now.

      And rents never rise as much as home prices…nor should they. Home prices incorporate an investment asset plus a consumption asset. So if rents always rise (for example) 20% as much as home prices, then obviously the spread will be larger the greater is the increase in home prices generally.

      So my prediction remains. OER is already at multi-year highs, but it’s going much higher (but “much higher” is 4% or 5% over the next year or two…not 18%!).


  1. No trackbacks yet.

Leave a Reply

%d bloggers like this: