What a low US mutual fund cash holding could mean

Mutual fund cash holding is usually the lowest at the peak of stock market indexes. So with only 3.6% of their asset in cash, could we be close to a top?

The following graph represents the S&P 500, in orange, versus cash holding, in white, inverted:

It definitely shows a sharp decrease from the “top” of March 2009, where 5.7% of mutual fund assets were in cash. It is a direct consequence of the ultra low return on cash, which gave strong incentives to money manager to “find” investments, and its corollary, a quick return of investors’ risk appetite.

This current level, 3.6%, has usually been associated with a high risk tolerance, which itself precedes major market tops. So in term of risk appetite, this indicator suggests a top could be close.

However, cash level seem to be trending lower since 1993, and we could imagine reaching a new low before a new top, especially since return on cash is litteraly zero.

The US labor market is improving. Isn’t it?

Some data have suggested the US labor market might be improving. For example the household survey, used to calculate the unemployment rate, has shown job creation last quarter, leading to a drop in unemployment rate to 9.7%. Weekly claims have also significantly decreased from the top of last year.

But other data are showing different picture. Yes the unemployment rate has decrease. But only on a seasonally adjusted (SA) basis. The non-seasonally adjusted (NSA) data are still showing an increase (at 10.6%):

Why should we look at the NSA data? Because the Bureau of Labor Statistics has to make assumptions about the seasonal impact. And as we discussed previously (link), the current crisis might have temporary altered some of those effects.

Secondly, the labor participation rate (ie. ratio between employed + unemployed and the overall size of the national population of the same age range) for men has significantly decreased:

Finally, the employment to population ratio (same ratio with only the employed) for men is at an historically low level:

Why does it matter? Because in a deleveraging economy, cash-flow is key to determine the speed of the adjustment. On the US households side, employment is the first source of income. So it is the variable to watch, especially if you believe there won’t be another asset bubble any time soon, propping up US household wealth.

US household total income vs. total debt

The New Normal, term coined by Mohamed El-Erian from Pimco, describes a new paradigm for the US economy. One of its aspects is the deleveraging of the US household. The main reason: 30 years of increasing debt burden, to finance the gap between their consumption and earnings, the former increasing much faster than the latter.

The next two graphs are showing this effect. The first one compares the increase in total income (in orange) versus the increase in total debt (in white) since 1980. Total income grew by +452% in nominal term, versus +1,031% for total debt:

And since 1952, for a long-term perspective. Total debt grew 3.5 times faster than total income:

As a consequence, households debt as a percentage of GDP grew significantly, especially in the last 10 years : special thanks to the housing “bubble” and home equity extraction. This is what you can see in the graph below, comparing nominal GDP versus total household debt (ratio debt to GDP at the bottom):

As you can see in all those graphs, this trend have paused. It could be only temporary, as some economists “hope”. We would then return to the old paradigm. Or it could be the beginning of a New Normal, where the American consumer deleverages. The reason I believe in this second narrative is because on the long-term, households have to match their debt growth with their cash-flow and asset growth.

And on the cash-flow side, it has not been the case for a long time. With a debt burden of more than 90% of GDP, it is hard to imagine how the US households will be able to add on more debt. Unless… the asset side of their balance sheet increases. But after a stock and a real estate bubble, it is hard to find a new asset class in their balance sheet that a) has not been hammered and b) that could have the potential for double digit annual returns.

So forget the American consumer, at least for now, he has to restore its financial ratios.

Chinese oil demand, where does the growth come from?

Where does the growth come from? Probably not from where you would expect it. According to Harry Tchilinguirian, senior oil analyst at BNP Paribas:

” China’s oil demand grew by 7.7% in 2009 according to IEA estimates but has been uneven across products. Gasoline’s growth in particular has been modest relative to booming sales of passenger vehicles. The top performance came in ‘other products’, which includes distillation residues like bitumen and coke, used in road building, and lubricants. Year-on-year growth in this category reached in excess of 350 kb/d, which is not surprising given China’s construction and infrastructure efforts. The other major growth product was naphtha, following large additions in capacity to China’s petrochemical industry. “

First surprise : modest growth in gasoline demand relative to the new car sales. Having a car there reflects a social status. But it is an inefficient mean of transportation in a those jammed city. So most new cars are barely used.

Second surprise: high growth in other products, which are used in construction and road building. This is probably another effect of the high level of fixed asset investments in China, especially in real estate and infrastructure. So if there is a slowdown in this sector, it would strongly impact China’s oil import growth. And probably for a long period of time.

Federal reserve balance sheet and bank reserves

In case you wondered were all the freshly printed money went, look at bank reserves. The following graph compares the FED balance sheet (in white) to the total bank reserves (in orange). The bottom graph calculates the ratio:

Most of the increase of the FED’s balance sheet is now sitting there, as bank reserve. With no increase in lending in sight, it probably will remain this way for a while. So far the FED has been successful at lifting asset prices, but has failed at channeling it into the economy. That is the limit of money creation, you can’t control were this new money will go.

Two scary graphs

I recently came across two graphs reflecting some of the underlying weaknesses of the US economy.

The first one charts the outstanding amount of asset-backed commercial papers. It is still deep diving:

As a reference, at the peak this market represented as much as 60% of the outstanding amount of all commercial papers:

The second one, in green, represents the number of auto sales in the US, adjusted for the change in population:

We are still very close to the bottom…

Consumer confidence

This morning te conference board consumer confidence number came lower than expected, at 46 vs. 55 expected (bloomberg survey).

“Stock market down, confidence down” I read somewhere… Probably. Those two data should have some level of correlation, as they are both related to the underlying strength of the US economy. Also, the wealth effect of increasing stock prices should have some impact on the American households confidence level:

Now the best correlation remains – very logically – with the US unemployment rates. The graph below charts the consumer confidence vs. the unemployment rate (inverted) :

Saudi Arabia’s oil exports

That is it, Saudi Arabia now exports more oil to China than the US :

The geopolitical consequences of such a shift are important.  Yes, Saudi Arabia remains under the US military shield, but its biggest customer is now China…

A “valuation” bear market for US stocks

A sideways stock market has some advantage : it improves valuation. If the S&P is pretty much sitting where it was 12 years ago, its valuation ratios have improved significantly. From a high of 5.27 for the price to book and 2.42 for the price to sales for example, they decreased to 2.17 and 1.23 respectively :

This is the real “bear-market”. Even during the 2003-2007 bull market, most ratios did not really expand. They rebounded from the “low” of 2002 and then range-bounded until late 2007. As a consequence, valuation did improved.

So value is in the process of coming back. We might not have seen the low, but we came a long way.

Also, as a side note, the ratio of book value to sales has steadily been increasing, from a low below 0.37 to now a high above 0.57. Which means companies’ book value has been increasing faster than sales. Margins have been trending higher in the last few year, which means higher earnings were extracted from the same amount of sales. As book value comes from retained earnings, we can assume that a large portion of this effect comes from those improving margins. Now, except from a short term bounce due to lay-off and cost cutting, I am not sure this trend will continue.

Bad “good ideas”

Single or double, and now even triple short ETF have become pretty popular. Advantage: they offer new ways to bet on the downside of multiple markets, not always easily accessible. So for small investors, they seem to be a good idea.

And they are… for short term trading. But not for long-term investment. For example, this is the S&P 500 versus SDS, the double short S&P ETF, since August 2007:

S&P 500 : down 25%, SDS : … down 22.5%! Almost as bad!

Worse, it is normal. It is due to the gamma of the ETF. Usually associated with options, this term is appropriate here : when you “normally” short a stock or an index, the variation in US$ of your exposure matched perfectly the variation of the underlying you are selling. For example, if you short the S&P 500 at 1,000, you are short $1,000 (to make it simple). Let’s say the S&P drops by 10% the next day. It’s value is 900, and your remaining exposure $900 (you also have $100 of unrealized profit). In this case, you have no gamma, your exposure varies exactly with the underlying price.

In the case of an inversed ETF and an index at 900, your ETF would be valued at $1,100 ($1,000 + 10% x 1,000). You are now short the S&P 500 at 900 for a net exposure of $1,100, not $900. This increase of $200 is due to the positive gamma, ie increasing short exposure when the underlying moves down. Said differently, you are now short 1.22 time the S&P, rather than 1.

In the case of the double short, the effect of the gamma is even bigger. With an index at 900, you now have an exposure of $1,000 + 2×10%*$1,000 = $1,200. So you increased your short exposure by $300 and are now short 1.33 time the S&P.

Now if the market rebounds by 10% the next day, your simple short ETF would be worth $1,100 – 10% x 1,100 = $990, while the index is only back to 990. So you would have lost 1% while the index would have lost 1%. With a double, you would have lost 2%. Repeat the same process (with smaller daily variations) for a few months and you have an almost certain loosing proposition, even if the underlying goes your way. On top of that, you are getting shorter and shorter as the market moves down, which is exactly at the wrong time, as the probability of a rebound (even just a technical one) increases.

So here again, just for the eyes, the long term effect of a large positive gamma associated with a short position (graph of UHPIX, the double short Hong Kong index, versus its benchmark – inversed) :

PS: This effect is also true with double or triple long ETF.