Checking in on the World's Best Rotation Strategy

Hello all... I knew I hadn't posted in a while, but I had no idea the last post was back in August. For those curious, this will not be the start of regular posting, though I have a few things I may share in the next few months.

Also... this definitely not the world's best strategy, but it seemed like the way to highlight it given the desire for investors to flock to excess confidence. In fact, there is no legit reason I can think of that explains why this seasonal pattern should outperform going forward, BUT the results (both in and out of sample) are good enough that I won't be ignoring them. 

For the better part of six years, I have provided an update on the "Secret Sauce" allocation strategy as soon as I start seeing all the "sell in may, go away" articles that always come this time of year. I didn't have to wait long.

To BusinessWeek:

Fine time to ask, then: Is an old-timey sell-off overdue, especially with “Sell in May and go away” season approaching?

Which brings me to the Secret Sauce, amazingly now in its 6th year of in sample testing.

What is the Secret Sauce?

An alternative to the "sell in May, go away" strategy, Secret Sauce is sell the S&P 500 in May and then invest in the Long Government / Credit bond index (rather than sit in cash). The "strategy" takes advantage of data mining that showed the Long Government / Credit index outperformed the equity market for the May through October time frame over the 34 years between 1974 and 2008.

The amazing thing is that since I first revealed the Secret Sauce back in July 2008, the results keep getting better. The massive outperformance since 1974 (14.9% vs 10.5% annualized returns) with lower volatility (12.3% vs 15.7% monthly standard deviation) was met by even larger outperformance over the past few years (21.3% vs 7.3% returns over the last 12 months, 17.3% vs 4.4% annualized over the last five) resulting in $1 invested in the Secret Sauce in January 1974 now being worth $227 vs. $1 in the S&P 500 being worth $50.5 (assuming no fees, transaction costs, or taxes).

Source: S&P / Barclays Capital

Employment Survey's Diverge

The Washington Post details:

One is called the payroll survey. It asks mostly large companies and government agencies how many people they employed during the month. This survey produces the number of jobs gained or lost. In July, the payroll survey showed that companies added 172,000 jobs, and federal, state and local governments cut 9,000.

The other is the household survey. Government workers ask whether the adults in a household have a job. Those who don’t are asked whether they’re looking for one. If they are, they’re considered unemployed. If they aren’t, they’re not considered part of the work force and aren’t counted as unemployed. The household survey produces each month’s unemployment rate.

In July, the household survey showed that the number of people who said they are unemployed rose by 45,000. In a work force of155 million, that doesn’t make a big statistical difference. But it was enough to raise the unemployment rate to 8.3 percent from 8.2 percent in June.

Unlike the payroll survey, the household survey captures farm workers, the self-employed and people who work for new companies. It also does a better job of capturing hiring by small businesses.

But the household survey is more volatile from month to month. The Labor Department surveys just 60,000 households, a small fraction of the more than 100 million U.S. households.

Looking at the data, the household survey was even worse than that. Not only did the number of unemployed rise by 45,000, the number leaving the workforce spiked (resulting in the number employed dropping significantly), resulting in the unemployment rate rising despite the better headline number.

Source: BLS

GDP Expands 1.5% in Q2

Bloomberg details:

The U.S. economy expanded at a slower pace in the second quarter as a softening job market prompted Americans to curb spending.

Gross domestic product, the value of all goods and services produced, rose at a 1.5 percent annual rate after a revised 2 percent gain in the prior quarter, Commerce Department figures showed today in Washington. The median forecast of economists surveyed by Bloomberg News called for a 1.4 percent increase. Household purchases, which account for about 70 percent of the world’s largest economy, grew at the slowest pace in a year.

Consumers are cutting back just as Europe’s debt crisis and looming U.S. tax-policy changes dent confidence, hurting sales at companies from United Parcel Service Inc. (UPS) to Procter & Gamble Co. (PG) Cooling growth makes it harder to reduce unemployment, helping explain why Federal Reserve Chairman Ben S. Bernanke has said policy makers stand ready with more stimulus if needed.

Looking at the data, we see the declining contribution by consumption, offset in part by a quarter over quarter (small) rise in investment and a less negative impact from government cuts (i.e. addition by the elimination of subtraction). Overall, considering what was going on during Q2 this report isn't awful, but certainly isn't encouraging considering we are now three years out of the recession.

Source: BEA

Leading Economic Indicators Decline by Most Since Last September

Bloomberg details:

The index of U.S. leading economic indicators fell more than forecast in June, a sign the U.S. economic expansion is slowing.

The Conferences Board’s gauge of the outlook for the next three to six months decreased 0.3 percent after a revised 0.4 percent increase in May, the New York-based group said today. Economists projected the gauge would drop by 0.1 percent, according to the median estimate in a Bloomberg News survey.

Source: Conference Board

Consumer Credit Jumps... a Good Thing?

Bloomberg details:

Consumer credit climbed more than forecast in May, led by the biggest jump in credit-card debt in almost five years that may signal Americans are struggling to make ends meet.

The $17.1 billion increase, exceeding the highest estimate of economists surveyed by Bloomberg News and the largest this year, followed a $9.95 billion gain the previous month that was more than previously estimated, the Federal Reserve said today in Washington. Revolving credit, which includes credit card spending, rose by $8 billion, the most since November 2007.

In a "normal" mean-reverting downturn, an increase in consumer borrowing is a good sign as it allows a consumer to maintain their purchase level (even without the current income to pay for it), with the expectation that they can cover their borrowing when their wages "revert" higher in the future.

A quote from someone who believes we are still living in the old world (we may be, I'm just not so certain):

“When the economy’s not doing well, that’s when you want the consumer to spend, and if it means borrowing to do that, then that certainly would be encouraged,” said Millan Mulraine, a senior U.S. strategist at TD Securities in New York, who projected credit would rise by $15 billion.

The concern is that we are now in year 4 year of the muddle through recovery, thus some concern that people are spending money on goods via debt that they may have trouble repaying. Back to Bloomberg:

A pickup in borrowing coincides with a slowdown in hiring and declines in consumer confidence that indicate the job market is failing to spur enough gains in wages to cover expenses. Employers added fewer workers to payrolls than forecast in June while the jobless rate stayed at 8.2 percent.

Regardless, overall nominal debt (even excluding student loans) is once again rising. A bullish sign for the short run. Mixed (in my opinion) for the longer run.

Source: Federal Reserve

Gaming the System... Disability Edition

In my post outlining the decline in employment (excluding teens), reader Mike pointed out another interesting phenomenon:

One of the stories I read about this month's BLS report dealt with the rise in people on disability. I thought charting this over a few dozen years and several recessions might tell an interesting story.

Unfortunately, the BLS only reports this data going back to 2008, but the most recent data shows an interesting trend. From the looks of it, we have what appears to be a not-so-coincidental mirror image between the decline in those unemployed with the rise in the number of people not working via claims of disability. I know next to nothing about disability, but my guess is individuals whose employment benefits have run out, have been increasingly taken advantage of disability (or tried to) as a replacement.

Source: BLS

Employment (Excluding Teen Hires) Turns Negative

Peter Boockvar provides details of the ugly job report:

June Payrolls totaled 80,000, 20,000 less than expected and well below the ADP whisper. The two prior months were revised down by a net 1000. The private sector added 84,000 jobs (13,000 from goods producing, 71,000 from services) vs expectations of a gain of 106,000. The unemployment rate held steady at 8.2% as the 128,000 increase in the household survey was basically offset by the 156,000 increase in the size of the labor force.

Below is an updated chart of the stagnant unemployment rate and slightly up ticking broader unemployment measure.

Unfortunately, things were even worse than that. When looking at the household survey, we see that the headline measure of unemployment doesn't account for the fact that teen employment (likely low pay part-time workers on summer break) accounted for more than 100% of all new jobs. Excluding teens (the second bracket from the left in the chart below), we can see that negative employment number. In addition, individuals over 20 continue to flee the workforce (more than 150,000 more 20+ year olds were classified as "not in the labor force").

A truly ugly report on first glance.

Source: BLS

Breaking Down Volatility of the VIX

I had never heard of the VVIX index (the volatility of the stock market's volatility "volatility of VIX") until a week back when Bill Luby from the always insightful 'VIX and More Blog' posted about the recent gap seen between the historical volatility "HV" of the VIX (i.e. the trailing volatility of the VIX index) and the implied volatility "IV" (i.e. what was currently priced in as the forward expected volatility of the VIX index).

To Bill:

Much to my surprise the current 20-day HV is 144, while the current IV is only 98. In other words, the markets expect the VIX to be considerably less volatile in the month ahead than it has been over the course of the last month.

But as we'll see, based on the historical relationship between the VIX and VVIX, the pricing wasn't all that unexpected. Over the previous month, the VIX index had fluctuated between roughly 17 and 26, so (ignoring any opinion of the market for a moment) the reading of 98 was actually right in the heart of where one would expect it based on historical data since the VVIX index 2007 inception.

And this is where things get interesting.

The historical chart shown above makes the assumption that the volatility of the VIX index is lower when the VIX index is lower, which would have been my first guess had I not really thought about it. When I did think about it, that actually makes no sense whatsoever.

Why?

• Volatility is a measure of change in percent terms
• A change from a low starting value is a much higher percentage change than an equivalent unit change from a high starting value (i.e. a 2 point change from 10 is a 20% move... a 2 point change from 80 is a 2.5% move)

With that understanding, the chart below showing the realized one month change from various VIX starting points shouldn't be too surprising because a 15 point move higher in the VIX from a starting value of 15 is much more likely than a 40 point move higher in the VIX from a starting point of 40.

The interesting thing is that this conflicts with how the VVIX (vol of VIX) has been priced. At low levels of VIX, the VIX is MORE likely to have wide outcomes (in percentage movement) while at high levels the mean reverting characteristic of the VIX has made a move down from high levels much more likely.

A chart attempting to summarize the first two charts is shown below. It shows the realized one-month forward volatility of VIX resulting from various starting VIX and VVIX index values. Interestingly enough, it shows that realized volatility of VIX spikes when the VIX index is low, but only when the market isn't pricing it in.

Potential thoughts from the above to consider further...

• When the VIX and the implied volatility of the VIX are low, buy options on the VIX (i.e. calls / puts)
• When the VIX is high and implied vol on the VIX is high, sell calls on the VIX

Source: CBOE

Common Investor Errors...

Early last week, Barry Ritholtz outlined what he believed were the top ten most common investor errors:

Here is my short list:

1. High Fees Are A Drag on Returns
2. Mutual Fund Are Inferior to ETFs
3. Reaching for Yield is Extremely Dangerous
4. Asset Allocation Decisions matter more than stock selection
5. Passive is usually better than Active Management
6. You must understand “The Long Cycle”
7. Behavioral Issues Are Costly
8. Cognitive Errors as well
9. Understand your own risk tolerance
10. Pay Guys Like Me For the Right Reason

While I think this is an interesting / solid list, I don't necessarily agree with a number of them. In his post I responded with the following:

Can we disagree on some of these?

2. Mutual Fund Are Inferior to ETFs: Too broad a statement. Some mutual funds are great (inexpensive, track indices almost exactly, prevent owners from day trading [see your #7] behavorial issues being costly), while many ETFs are very poor (broad tracking error, expensive, leveraged inverse ETFs)

3. Reaching for Yield is Extremely Dangerous: Everything is based on appropriate compensation for the risk an investor takes… 5 years ago an investor sitting in cash received [a 4-5%] risk-free return. 0% [yielding] cash is now return-free risk. An investor “reaching for yield” now may actually now be a risk reduction exercise.

4. Asset Allocation Decisions matter more than stock selection: agree, but by definition asset allocation decisions are “active” decisions, hence….

5. Passive is usually better than Active Management: seems in conflict with #4

Diving right into my point #2 (because ETFs seem to be uniformly praised these days) is that mutual funds are not all inferior to ETFs (stating ETFs are superior is too broad of a statement). This is especially true for sectors / asset classes where the underlying securities are less liquid and the ETF itself trades with minimal volume (volume isn't nearly as important if the underlying securities are liquid... a point for another day). In these instances, it is more likely that the price of the ETF can move significantly from the ETF's net asset value "NAV" (the actual value of its holdings) and the bid/ask spreads widen, both of which can be negative to an investor.

One example can be seen year-to-date with Muni ETF MUB performance relative to that of a muni mutual fund. I am not vouching for the Fidelity fund below (it was the first to come up when I looked for a national muni fund with roughly 7 years of duration).

In addition to the underperformance of MUB, the volatility is 3x higher at 6.8% vs 2.2% due to the widely fluctuating ETF price vs the underlying NAV which hit a 4% premium in February (i.e. someone buying that day paid 4% more than the securities were worth).

Source: Yahoo Finance

(Almost) All Assets up in the First Half

Despite a volatile second quarter, risk assets (actually all assets with the exception of commodities) performed quite well in the first half. Leading the pack were REITs (up both quarters) on demand for income, a potential inflation hedge, and CHEAP real estate financing.

Source: Yahoo

More on Corporate Profits

Back in September of last year, I showed this chart outlining that corporate profits as a percent of GDP were approaching a three-standard deviation event. Since that time, the relationships has gotten even more extended hitting an all-time high.

Which brings me to this morning's tweet from PIMCO's Bill Gross:

Simple formula: US profit growth rate = (real GDP x 5) – 10. No “ka-ching” at 2% or less GDP growth.

I appreciate the insight that profits can be thought of as being leveraged to economic growth (hence the wide fluctuations), but struggling to see why nominal profit growth would have a relationship with a real (after inflation) economic growth or why those specific numbers (why 2% real and not 2.5% real?) were used.

Anyhow... I was interested enough to see what this equation looked like using actual data, so I put together the below chart going back 60 years against corporate profits, as well as against my old simple standby of using nominal economic growth (my preferred long-term measure for corporate profits as simple math tells you that corporate profits can't grow faster than the economy over the long term or else they'd be bigger than the economy itself - hence profit growth tends to mean revert relative to nominal growth).

The chart...

What do we see?

Well corporate profits are basically right on trend (a surprise to me), but nominal growth is well below trend and the PIMCO formula (the formula based on 2% real growth) is WAY below trend, indicating corporate profits are significantly above trend (by that record 30% level relative to nominal growth and an off the charts 200% relative to the PIMCO formula).

Seems like the old nominal GDP standby has historically been more reliable, but I will be thinking more about the insight that corporate profits are leveraged to (and in need of) specific levels of economic growth. If anything, this may indicate earnings are potentially more stretched than I previously thought.

For those interested, there was a great piece by GMO on the topic of extended profits a few months back.

Source: BEA