This will be a very short remark, partly because I am certain that someone else must have observed this already.
The Dow Jones Industrial Average declined on Tuesday after having risen in each of the preceding 12 days. I was curious, and the DJIA has data going back more than a century (unlike, for example, the S&P 500, the Russell, or other indices), so I checked to see how often that has happened before.
It turns out that only three times before in history has the Dow advanced in 12 consecutive sessions. The dates of those occurrences are (listed is the last day of advance before the first decline):
July 8, 1929
December 7, 1970
January 20, 1987
The latter of these three was actually a 13-day advance, and the longest in history.
Now, the 1970 occurrence seems to be nothing special. It occurred five years into a 15-year period that saw the Dow go nowhere in nominal terms, but there was nothing special about 1971. However, anyone who invests in the stock market ought to know the significance of 1929 and 1987. It also bears noting that current market valuations are higher (in terms of the Cyclically-Adjusted PE ratio) than on any of those three days – quite a bit higher, in fact.
None of which is to say that we won’t have another 10, 20, or 30-day streak ahead of us. I suspect the bulls will say “see? This same occurrence in 1929 and 1987 happened months before the denouement. We still have time to party!” And they may be right. This isn’t predictive. But it, especially when compared to valuation levels second only to those seen at the peak of the “Internet Bubble,” is ominous. This is a party I wouldn’t mind missing.
Walmart (WMT) didn’t have its best day today. The bellwether retailer forecast a profit decline of 6-12% in its 2017 fiscal year, in some part because of a $1.5bln increase in wage expenses; the stock dropped 10% to its lowest level since 2012 and off about 33% from the highs (see chart, source Bloomberg).
I mention Walmart neither to recommend it nor to pan it, but only because in the absence of news from WMT I would have been inclined to ignore the modest downside surprise in Retail Sales today; September Retail Sales ex-auto-and-gasoline were unchanged versus expectations for a +0.3% rise. But Retail Sales, like Durable Goods, is a wildly volatile number (see chart, source Bloomberg).
This was a bad month, but it wasn’t the worst month in 2015. It wasn’t even the second or third-worst month in 2015. Looking at a monthly figure, it is difficult to reject any null hypothesis; put another way, you really cannot discern whether +0.5% is statistically different from +0.0%. [I didn’t actually do the test…I am just making the general statistical observation.] Today’s data will tweak the Q3 forecasts a bit lower, but isn’t anything to be upset about. Except, that is, for the fact that Walmart is bleeding.
There is something else that is different about this decline, and really about this whole year. I have documented in the past the steady decline in equity volumes that has been occurring for almost a decade now. The chart below shows the cumulative NYSE volume, by trading day of the year, for 2006 through present. Note the steady march lower in volumes year after year after year. 2014 and 2013 were almost mirror images, so you can’t see 2014. But notice the thicker black line: that is 2015.
|Number of sub-billion share days|
In 2015, we are on pace for a mere 228 sub-billion share days.
I guess by now my point is plain, but here is one more chart and that is the rolling 20-day composite volume for 2014 (lower line) and 2015 (upper line).
In general, volumes have been higher this year, but the real divergence began at the end of July, when the lines began to move away from each other more rapidly. The equity breakdown started on August 20th.
What does this all mean? Rising trading volumes while markets are declining suggests we should consider imputing more significance to what many are calling a correction but which may be the beginning of something deeper. There are re-allocations happening, and outright sales – not just fast money slinging positions around. Technically, this is supposed to put more weight on the “damage” done by this correction, and raise a bit of a warning flag about the medium-term set-up.
Incidentally, you can buy warning flags cheaply at Walmart.
In an excellent (and free!) daily email I receive, the Daily Shot, I ran across a chart that touched off my quant BS alert.
This chart is from here, and is obviously a few years out-of-date, but that isn’t the problem. The problem is that the chart suggests that gold prices rise 5.5% every year. If you buy gold in January, at an index value of 100, and hold it through the flat part of January-June, then you reap the 5% rally in the second half of the year.
No wonder people love gold! You can get a 10% annual return simply by buying in July and selling in December!
The problem is that this is not the way you should do a seasonal chart. It has not be detrended. We detrend data because that way, we can express the expected return for any given day as (the normal expected return) plus (the seasonal component). This is valuable because, as analysts, we might have a general forecast for gold but we will want to adjust that forecast to a holding period return based on a seasonal pattern. This is very important, for example, with TIPS yields and breakevens, because inflation itself is highly seasonal.
Now, the seasonal chart done correctly still suggests that the best time to own gold is in the second half of the year, but it no longer suggests that owning gold is an automatic winner. (It is a separate argument whether we can reject the null hypothesis of zero seasonality altogether, but that’s not my point here).
Frankly, I would also use real prices rather than nominal prices, since it is much easier to make a statement about the expected real return to gold (roughly zero over time, although it may be more or less than that based on current valuation metrics) than it is to make a statement about the expected nominal return to gold, since the latter includes an embedded assumption about the inflation rate, which I would prefer to strip out. And I would also include data from the 1970s.
The stock market, the bond market, the commodities markets (to a lesser extent), FX markets – they are all experiencing a marked increase in volatility.
Some observers want to call this bearish for equities, mainly because they already are bearish. This is a very bad reason. While really bad equity returns almost always occur coincident with a rise in volatility – the old maxim is that stocks go ‘up on the staircase and down on the escalator’ – that does not mean that volatility causes bad returns. Or, put another way, there are also periods of increased volatility that do not precede and are not coincident with bad returns.
However, there actually is a reason that increased volatility might lead to poor short- to medium-term returns, that isn’t based on technical analysis or spurious correlations. Moreover, a relatively new phenomenon (the rise of so-called ‘risk-parity’ strategies) is starting to institutionalize what was already a somewhat natural response to volatility.
In ‘risk-parity’ strategies, the weight of an asset class (or a security within an asset class, sometimes) is inversely proportional to the risk it adds to the portfolio. Generally speaking, “risk” here is defined as variance, because it is easy to estimate and there are markets where symmetrical variance trades – i.e., options markets. But what this means is that when volatility (sometimes realized volatility, and sometimes option “implied” volatility) rises in stocks, then risk parity strategies tend to be shedding equities because they look riskier, and vice-versa. Right now, risk parity strategies are likely to be overweight equities because of the long period of low realized and implied volatility (even though the valuation measures imply quite high risk in the sense most of us mean it, in terms of the probability of return shortfall). Risk parity strategies are probably superior to ‘return-chasing’ methodologies, but by being ‘risk-chasing’ they end up doing something fairly similar when they are all operating together.
Note that while risk-parity strategies are comparatively new – well, not exactly because it is an oldish idea, but they have only recently become a big fad – this general phenomenon is not. The natural response to greater equity market volatility is to pare back exposure; when your broker statement starts to swing around wildly it makes you nervous and so you may start to take some profits. This is also true of other asset classes but it seems to me to be especially true in equities. Nobody who gets involved in commodities is surprised at volatility: the asset class suffers from a midguided belief that it is terribly volatile even though commodity indices are just about exactly as volatile as equity indices over time. But equity investors, contrariwise, seem perennially surprised at 2% moves.
So, while the recent volatility doesn’t mean that a move lower in equities is assured, it increases the probability of such because risk-parity strategies (and other investors reacting nervously to overweights in their equity exposure) will begin to scale back positions in the asset class in favor of positions in other asset classes, probably mostly bonds and commodities. At this point it would be good for me to point out that only the very short-term volatility measures have moved up dramatically; the VIX is well off its bottom but only up to 18.8 and it has been there numerous times in the last few years (see chart, source Bloomberg). But the longer the volatility continues like we have seen it for the last week or two, the bigger the chances that the asset-allocation boxes start to make important shifts (and the quant hedge fund boxes will probably move a bit before those asset allocation boxes do).
As an aside, the tendency for asset allocation shifts to follow volatility shifts is not the reason that the VIX displays a strong inverse directionality. Neither is the main reason for this inverse directionality because the VIX is a “fear gauge.” The main reason is that the VIX weights near-the-money options more heavily than out-of-the-money options. Because options skews almost always imply more downside volatility for stocks than upside volatility, when the market declines it tends to bring more “high volatility” strikes into play and so part of the VIX increase in a down market is simply mechanical.
I am not calling for a sharp decline in stocks, nor for an extended decline in stocks. My position and view is as it has long been, that the prospect for attractive real returns from equities over the next 5-10 years is quite small and beaten handily by commodities’ prospective returns at that end of the risk spectrum. I don’t think that most investors (me included!) should swing asset allocations around frequently in response to technical indicators or such things as “momentum”, but rather should focus on evaluating expected long-term returns (which are somewhat predictable) and invest for value. And I must admit I also think that “risk-parity” is a clever marketing gimmick but a pretty absurd way to assemble a portfolio for almost everyone. My point here is to highlight one little-considered aspect of herd behavior, and how that herd behavior may have become more institutionalized as late, and to consider the risks that herd behavior may create.
 This in turn is not due so much from the tendency of markets to have more downside volatility than upside volatility, but from the fact that buying protective puts and selling “covered” calls are both considered “conservative” options strategies. So, out-of-the-money puts tend to be too expensive and out-of-the-money calls too rich.
As we head into a very busy week of economic data, the bond market remains drippy with the 10-year yield up to 2.59%. (Just writing that makes me laugh. Who would have thought, only a few years ago, that 2.59% was a high-ish yield?)
How we got here, from the ultra-low levels of the last two years, is well-traveled territory. The Fed’s swing from “QE-infinity” to “someday, maybe, we might not buy as many bonds” helped trigger a run for the exits, and then negative convexity inflection points kept the rout going for a long time. Most lately, the threat of muni bond convexity has been looming as the next big concern.
But my message today is actually one of good cheer. The worst of the bond selloff was now more than three weeks ago, without a further low being established. In my experience, convexity-inspired selloffs typically end not with a sharp rebound but with a sideways trade as “trapped” long positions gradually work their way out and buyers start to nibble. But it remains a buyer’s market for several weeks, at least.
We are getting far enough along in that process that I suspect we have a rally due. This has nothing to do with any economic data coming up. There is enough data coming this week, from Consumer Confidence to Payrolls to GDP to the Fed statement, that both bulls and bears will be able to find something to point to. And I am not pointing to technicals, exactly. I am just saying that markets rarely move in a straight line, and even bear markets – such as the one I think we have now entered, in bonds – have nice rallies from time to time.
But here’s a reason to expect this to happen relatively soon. The chart below is a neat “seasonal heat map” chart from Bloomberg showing the monthly yield change for the last 10 years and the average monthly change on the top line.
For a long time, I have been following the rule of thumb I learned as a mere babe in the bond market, and that’s that the best time of the year to buy bonds is the first few days of September. From at least the late 1970s until today, September until mid-October has been the strongest seasonal period of the year (not every year, but with enough consistency that you wanted to avoid being short in September). But the heat map above shows that this tendency may have shifted. The month that has seen the best average bond market performance over the last decade has been August, with yields falling an average of 22bps with rallies in 8 of the last 10 years. If we were sitting with 10-year yields at 1.59%, I would be less interested in this observation, but at 2.59% I am looking for the counter-trade.
To be sure, yields in the big picture are headed higher, not lower. But I am looking for signs that the recent selloff has over-discounted the immediate threat of ebbing Federal Reserve purchases. And I don’t expect growth to suddenly leap forward here, either.
As an aside, 10-year TIPS yields have also experienced one of their best months in August, with the other clear positive month being January. But, because nominal yields have been so strong, August has been the worst month for breakevens, with 10-year breakevens falling 10bps on average over the last ten years. No other month has seen breakevens decline as much as 6bps, on average.
Now, although I am a bond bear in the big picture, I don’t think that the housing market is doomed because interest rates will go up one or two or three percent. I am fascinated by how many analysts seem to think that unless 10-year rates are below 3%, the housing market will collapse. I argued about six weeks ago that higher mortgage rates should not impact sales of homes very much as long as the interest rate is less than the expected capital gain the homeowner expects to make on the home. (Higher rates will, however, cut fairly quickly into speculative building activity, which is much more rates-sensitive). And here is another reason not to worry too much about the housing market. A story in Bloomberg last week says that adjustable-rate mortgages are booming again, with mortgagees taking them out at the highest pace since 2008. Faced with higher rates, and a Fed with is not likely to raise short rates for a long while – as they have taken pains to keep reminding us – homebuyers have rationally decided to take the cheaper money and let the future refinancing take care of itself.
Whether that is sowing the seeds of a future debacle I will leave to other pundits to debate. From my perspective, the important point is that higher rates are not likely to slow home sales, or the recent rise in home prices, very much…unless they get a lot higher.
The beatings are continuing, and apparently morale really does improve with such treatment. Consumer Confidence for June vaulted to the highest level since early 2008, at 81.4 handily beating the 75.1 consensus. Both “present situation” and “expectations” advanced markedly, although the “Jobs Hard to Get” subindex barely budged. It is unclear what caused the sharp increase, since gasoline prices (one of the key drivers, along with employment) also didn’t move much and equity prices had been steadily gaining for some time. It may be that the rise in home prices is finally lifting the spirits of consumers, or it may be that credit is finally trickling down to the average consumer.
Whatever the cause, it is not likely to prevent the rise in money velocity that is likely under way, driven by the rise in interest rates. Between the rise in home prices – the Case-Shiller home price index rose a bubble-like 12.05% over the year ended April, and Existing Home Sales median prices have advanced a remarkable 14.1% faster than core inflation (a near record, as the chart below shows) over the year ended in May. (Lagged 18 months, such a performance suggests about a 3.9% rise in Owners’ Equivalent Rent for 2014).
The nonsense about deflation is incredible to me. Euro M2 growth hasn’t been this high (4.73% for year ended April) since August of 2009. Japanese M2 growth hasn’t been this rapid (3.4% for year ended May) since May 2002. US money supply is “only” growing at 6.5% or so, down from its highs but still far too fast for a sluggishly-growing economy to avoid inflation unless velocity continues to decline. But you don’t have to be a monetarist to be concerned about these things. You only need to be able to see home prices.
Core inflation in the US is being held down by core goods, as I have recently noted. In particular, CPI for Medical Care just recorded its lowest year-on-year rise since 1972, and Prescription Drugs (1.32% of CPI and an important part of core goods) declined on a y/y basis for the first time since 1973. The chart below (source: Bloomberg) illustrates that as recently as last August, that category was rising at a 4.0% pace.
Now, I suspect that this has something to do with Obamacare, but no one seems to know the full impact of the law. Keep in mind that Medical Care in CPI excludes government spending on medical care. So, one possible narrative is that the really sick people are leaving for Obamacare while the healthy people are continuing to consume non-governmental health care services. This would be a composition effect and would imply that we should start looking at CPI ex-medical for a cleaner view of general price trends. I have no idea if this is what is happening, but I am skeptical that prescription meds are about to decline in price for an extended period of time!
But that’s the bet: either core inflation is going to go up, driven by things like housing, or it’s going to go down, driven by things like prescription medication. Place your bets.
Equity prices recovered today, but bond prices continued to slide into the long, dark night. For a really incredible picture, look at the chart below (source: Bloomberg), which shows the multi-decade decline in 10-year yields on a log scale, culminating in the celebrated breakout below that channel. Incredibly, the recent selloff has yields back to the midpoint of the channel and not outrageously far from a breakout on the other side!
Incidentally, students of bond market history may be interested to know that the selloff has now reached the status of the worst ever bond market selloff (of 90 days or less) in percentage terms. Since May 2nd, 10-year yields have risen from 1.626% to 2.609%, a 98.3bp selloff which means that yields have risen 60.5% in less than two months.
And we are probably not done yet. I wrote about a month ago about the “convexity trade,” and I made the seemingly absurd remark that “This means the bond market is very vulnerable to a convexity trade to higher yields, especially once the ball gets rolling. The recent move to new high yields for the last 12 months could trigger such a phenomenon. If it does, then we will see 10-year note rates above 3% in fairly short order.”[emphasis in original] Incredibly, here we are with 10-year yields at 2.61%, up 60bps over the last month, and that statement doesn’t seem quite so crazy. As I said: I have seen it before! And indeed, the convexity trade is partly to blame for what we are seeing. I asked one old colleague today about convexity selling, and here was his response:
“massive – the REITs are forced deleveraging and there are other forced hands as well. The real money guys are too large and haven’t even sold yet – no liquidity for them. The muni market has basically crashed and at 5% yields in muni there is huge extension risk on a large amount of bonds: something like $750bln in bonds go from 10-year to 30-year maturities as you cross 5%.” (name withheld)
Now, I am not a muni expert so I have no idea what index it is I am waiting to see cross 5%. But the convexity trade is indeed happening.
Lots of bad things have happened to the market, but they really aren’t big bad things. In fact, I move that we stop using the term “perfect storm” to mean “modestly bad luck, but I had a lot of leverage.” The Fed was never going to be aggressively easy forever, and as various speakers have pointed out recently they didn’t exactly promise to be aggressively tightening any time soon. There is bad news on the inflation front, but the market is clearly not reacting to that. Some ETFs have had some liquidity issues, and emerging markets have tumbled, and there was a liquidity squeeze in China. But these are hardly end-of-the-world developments. What makes this a really bad month is the excess leverage, combined with the diminished risk appetite among primary dealers who have been warned against taking too much “proprietary risk.”
And markets are mispriced. Three-year inflation swaps imply that core inflation will be only 1.9% compounded for the next three years (the 1-year swap implied 1.6%; the 2y implies 1.75%). That is more than a little bit silly. While I have not been amazed that the convexity trade drove yields very high, and probably will drive them higher, it has surprised me that inflation swaps and inflation breakevens have continued to decline. Still, investors who paid heed to our admonition to be long breakevens rather than TIPS have done quite a bit better, as the chart below (source Bloomberg), normalized to February 25th (the date of one of our quarterly outlook pieces) illustrates.
As the bond selloff extends, I don’t think TIPS will continue to underperform nominal bonds. I believe breakevens, already at low levels (the 10-year breakeven, at 1.97%, is lower than any actual 10-year inflation experience since 1958-1968), will be hard to push much lower, especially in a rising-yield environment.
It’s hard for me to truly grasp the reality of a world in which the downgrade of the British Empire’s credit (late on Friday) was the third most-important story, but so it is.
The UK was dropped from AAA to AA1 (one notch, but an important one) by Moody’s on Friday, and sterling dropped to the worst level against the dollar since 2010. In the grand scheme of things the drop to $1.51 was not critical, and the cable is still almost in the range it has held for the last few years, but some technicians are sure to see the breakdown as an ugly technical development (see chart, source Bloomberg).
But, fortunately for Britain, the Italians were drawing global attention to themselves and the Euro. As ballots were counted in the election to establish the balance of power in that nation, global markets careened up and down depending on the latest tallies. Ultimately, it appeared that a split government was in the offing, with a general repudiation of the politicians which have been party to austerity measures. The party of Berlusconi, who ran opposing the austerity measures, combined with the “Five Star Movement” party of Grillo, who advocates suspending interest payments on Italian debt and holding a referendum on Italian membership in the Euro, would represent an outright majority in the Senate although the lower house ends up in the hands of Bersani because of a “bonus premium” that guarantees the winning coalition will have a majority.
In the end, the reason the Italian election matters more than the downgrade of the UK isn’t because the election raises questions about whether Italy is committed to austerity; it’s that the election raises questions about whether Italy is committed to the Euro. This isn’t Greece. With a $2 trillion economy, Italy is the third largest member of the Eurozone, behind Germany ($3.4T) and France ($2.6T). It is the size of the other four PIIGS combined. And they’ve also issued a lot of inflation-linked bonds, by the way, so look carefully if you own an inflation-linked bond fund that invests in non-US bonds, just so you know.
Now, Italy isn’t going to default any time soon. They’re going to have another election, and in the lead-up to that one there will be more concern and angst. But then the leaders will use that as a bargaining chip, etc. etc.. We’re a long way from a default or exit of Italy from the Euro. But we’re probably not as far from fear of default or exit.
Still, the immediate uncertainty is past. The markets will calm back down reasonably quickly (which doesn’t mean they’ll rally, being overpriced to begin with). Each successive fire drill will cause a shorter and less-intense period of instability in Europe, until eventually the crisis completely passes, or one episode turns out to be qualitatively different and the whole thing breaks down.
And speaking of episodic crises brings us to fiscal cliff redux. The U.S. will hit the sequester barrier in a few days, with almost no chance that it will be averted. The Republicans seem comfortable that this isn’t such a big deal, and that if it turns out they are right then the scare tactic they feel is being used against them will be defanged. The Democrats seem to believe (and intent on making sure everyone else believes) that any cut in expenditures is tantamount to the End of Days. I don’t think the market ought to react very seriously to it, because we’re only talking 0.25% of GDP, but that all depends on how much hyperventilating we get from the media.
Still, it’s an interesting story because if it turns out that the budget can be cut by 2% (albeit 2% from baseline, which is still an increase over last year) without the economy going into the loo, then we’ve moved the goalposts for future negotiations. And if both sides can understand that, then cutting spending (even real spending!) by 2% per year will slowly get the budget back on a course that, while not sustainable, at least doesn’t lead to immediate immolation.
I am not sure how stocks will react to all of this (have I mentioned they seem expensive?), but I know that all three stories should be bond-bullish. The 10-year yield made it all the way back to 1.87% today after peeking over 2% several times the last few weeks. I think there is further upside to bonds for now, and that may mean that breakevens can also retreat some from near all-time highs. If I am right, then selling 10yr notes if they approach 1.65% or buying 10-year BEI near 2.40% represent better placement for the long term trades, which I expect to be higher in yield and in breakevens over 2013.