May 19, 2007: Seasons in the Sun

US small-caps under-performing:
After 7 years of small-cap outperformance, US stock markets are starting to show some large-cap bias. What's the evidence? Why does it matter? A lot of nice charts today. Enjoy.

Yet another new DJI, (and soon S&P500,) all-time record

A bit over 7 years ago, in March-April 2000, US markets reached a peak that was going to stay in place for a long time to come. In the following 2000-2002 bear market, the technology ladden NASDAQ which led the way up, had lost about 70% of its value in about 2.5 years and the S&P500 about 50% of its value in the same period.

The US markets have bottomed in the fall of 2002 and have been going up for nearly 5 years. What may be a bit more interesting is the time it took different asset classes to recover.

If you listen to the news, the headlines keep telling the story of new all time highs for the DJI (Dow Jones Industrials Index). Every time the Dow adds a 1000 points it seems to generate a big story.

Doing the math makes it obvious that going from 10,000 to 11,000 (+10%) is not the same as going from 12,000 to 13,000 (+8.3%). In other words, every 1000 points as the Dow goes up should get less and less of a big headline. I don't think they'll stop though. Big round numbers make good headlines.

The S&P500 looks likely to finally join the company of the Dow Jones sometime soon. This will no doubt generate more big headlines.

What about the NASDAQ? It is still way off its high. Rest assured, it will not get there any time soon. Those who bought technology in 2000 will have to wait some more years to recover.

Different indexes are very different from each other.

The (now) forgotten real winners: Russell 2000 & 3000, small caps

Looking beyond the DJI and S&P500 indexes brings much more useful insights.

After the 2000-2002 bear market, the fastest asset-class to recover were the small stocks and especially small-cap value stocks. The Russell 3000 and the Russell 2000 didn't wait for the Dow or the S&P500 to make new highs. They've done it almost 4 years earlier. Those who were invested in small-cap value stocks have been celebrating new highs for 4 years now, and they weren't hit nearly as hard as the NASDAQ investors to begin with.

Yes, small cap and value stocks have been outperforming the major indexes (which are large-cap by definition) for 7 years now. Starting from the April 2000 peak, throughout the bear market where they declined much less than their large-cap counterparts, and the following bull when they continued to greatly outperform. Small caps have been on a long tear.

While the NASDAQ is still a long way off its April 2000 peak, the Russell 2000 (represented by the ETF: IWM) small cap index, has broken its all time high way back in 2003.

One remarkable and little known fact is that small-caps actually had a positive year in 2001. They have gained 2.5% during the year in the middle of one of the biggest bear markets in US history.

Here's a chart showing how market leadership has changed between 1993 and 2002 among different asset classes. The best performer in each year is at the top row and the worst in the bottom row.

What I like the most about the above chart is that it highlights my investing philosophy. In order to do well over the long term, you don't need to pick the very best asset-class every time. You should be happy to be no.2 or no 3. In fact: just by avoiding the very worst of all (like large-cap growth in most of the 2000-2002 period), you can do pretty well.

This is what this article is all about. Change of seasons, and how to avoid the worst with some additional margin of safety.

One of my favorite "all season" comparison charts

Part of the recent 7 year outperformance of US small caps can be seen in one of my favorite all-season charts. It is a chart I look at periodically, just to get a feel for the current market preferences.

It compares 5 ETFs by their average maket-cap since it is convenient to simply use ETFs designed to replicate the different indexes. We compare 5 ETFs from small-cap (top) to large-cap (bottom) over the past 5 years. All these ETFs are market-cap weighted. VTI (center blue line and title of chart) is 'special' in that it is used as a reference point for the others:


ETF	Color	Description
---	-----	-----------------------------------------------------------
IJR	Orange	iShares S&P 600 Small-Cap stock Index, just smallcaps
MDY	Black	MidCap SPDRs, S&P 400 Mid-Cap stocks
VTI	Blue	Vanguard total US market index: ~5000 stocks (both small & large)
SPY	Green	S&P500: top 500 large-cap US companies
OEF	Red	S&P100: top 100 largest cap US companies (top subset of SPY)

What the chart tells me is that some market preferences last for a long time. You don't need to pick exact relative bottoms or tops, which are impossible IMHO, in order to ride the winners. You can be a bit late to the party.

Another useful lesson is that it is much easier to understand the markets using comparative approaches rather than absolute observations. For example when something you own drops, it may look bad, but if it drops less than the market it means you may actually be doing pretty well. The lines in the chart above go up and down in a seemingly random fashion, yet the top line keeps gaining more and more compared to the bottom line.

If you look at the right end of the chart, you can see that some of the divergence has been reversing. One way to see this is to notice that the red bottom line (OEF) took a shorter time to recover from the recent February-March correction than the top (orange) line. But there's an easier way: just look at the past 6 months instead of the past 5 years:

Here we see IJR (small-cap) clearly losing altitude in relative terms. While midcap (MDY) is still hanging in the lead and SPY (S&P500) is catching up with VTI (the total market tracking ETF).

Is it time to underweight small? I think so. It is also important to realize that Value still reigns over growth so overweighting large cap value (IWD) over large-cap in general (SPY, OEF) seems even better than merely switching to the S&P 500 index right now. This too may eventually change.

The seemingly random leadership cycle

Leadership in the markets is a cyclical and seemingly random thing. There's no simple formula telling us in advance whether large-cap or small-cap will outperform, or whether value stocks would outperform growth stocks. But there are some rules that can help us. The first one obviously is statistics and history. Finance professors Eugene Fama & Kenneth French have shown in what is perhaps the most cited finance papers of all times, that:

During the period 1950 to 2000 (51 years) leadership has cycled among all 4 basic asset-classes, but the portion of time where each of the asset-classes led the market was not evenly divided. The worst performer was small-cap growth which led only 8.5% of the time. The best was small-cap value which led 42.6% of the time. The 4 portions were:


	Asset-Class		Leading % of time
	---------------		-----------------
	LargeCap Growth		17.0%
	LargeCap Value		31.9%
	SmallCap Growth		 8.5%
	SmallCap Value		42.6%

There are good and strong reasons for this uneven performance persisting for so many years despite being such a widely known fact. Fama & French themselves have a theory called "The 3 factor model". In my humble opinion, a better and much more intuitive explanation was given by moneychimp here. Pick whichever you like. I believe the outperformance of small-cap value in the very long term (say during every 20 year period) will continue in developed markets (emerging markets seem different, they seem to prefer large-cap). It is caused by market psychology, market maturity and transparency, and human nature.

So betting on the value half, means winning almost 75% of the time. Betting on size (small vs large) is, based on history, a pretty even 50%/50% bet.

We could simply put our tokens on the value half of the market (spread about evenly among all market-cap sizes, with perhaps a small bias towards small-caps) and just forget about the rest. Assuming history will repeat itself, we should win 75% of the time.

But we can do better: any long time observer of the markets knows that long time leaders eventually change. Anything that has outperformed for many years, is less and less likely to continue to outperform. Since small-cap value had a very long run, it may be time to underweight it for a while after it ran its course. Can we do better than 75% of the time? I think we can. While large vs. small seems to be a 50%/50% bet, it gets easier after very long runs of either one.

Here is a nice set of charts from DFA (Dimensional Fund Advisors) showing the size (large vs small) market bias during different times in history. It covers the 1927-2000 (73 years) period.

Again, what we see here is that these periods of outperformance take years, sometimes 8 years until they run their course. After 7 years, we are now at or near the end of a very long run in small caps. Just like investing in large-growth was the worst thing one could do at the 2000 peak, overweighting small-cap after 7 years of leadership is beginning to look questionable.

A maturing bull market

In the past 4 years, the US market leadership consisted of many relatively unknown names (small caps). A sign of a mature bull market is large caps (well known names) outperforming with an narrowing of leadership over time. That's usually when peaks occur. We're not yet there but if indeed smallcaps have stopped leading, we are getting closer.

A good indicator for bull market maturity is the quarterly comparison of active mutual funds with the major indexes. As you may know, most mutual funds and active managers fail to outperform the indexes. Anyone who has to manage many billions in assets, would find it extremely hard to outperform the index especially during a bull market and most of all as they get closer to their peak. In the past few years beating the indexes has been relatively easy by overweighting foreign, commodities and small-caps. With small-caps losing leadership, it is getting more difficult.

Sure enough, we got that signal recently:

INDEXES OUTPERFORMED ACTIVE FUNDS IN FIRST QUARTER
Standard & Poor's reported that the returns of the major indexes led active funds in both the small and mid-cap categories in the first quarter of 2007, while large-cap actively managed funds did beat the S&P 500 in the period. The S&P Indices Versus Active Funds Scorecard (SPIVA) results, showed that in the first quarter the S&P MidCap 400 outpaced 84.8% of mid-cap funds and the S&P SmallCap 600 led 53.5% of small-cap funds; 64% of their large-cap brethren beat the S&P 500.

What you see here is similar to what we saw in the 6-month charts: small-caps are (relatively) the weakest compared to the active managers. Mid caps are the strongest, and large-caps are finally starting to gain ground. Comparative analysis is a very useful tool.

This tells me again that it is time to move up from small to large caps while, at this point at least, keep overweighting value. A mix of IWS (iShares midcap value), or VOE (Vanguard midcap value) with the larger cap counterparts: IWD (iShares large cap value) and VTV (Vanguard Value, which is almost identical) is IMHO the way to go in the US for the rest of the year.

What the US dollar says

The US dollar has recently hit a low of 2 USD per 1 British Pound Sterling. It keeps hiting lows against other currencies, even against the Israeli Shekel. While I don't see much on the horizon that can support the dollar when trade deficits are so high and when countries with surplus dollars (e.g. China) want to diversify out of dollars, I do see a great benefit for the large-cap US firms exporting products and services and globalizing. This is another boon for large-cap US stocks.

What Libby says

Libby who? you ask. Well, I just stumbled upon a money manager from Livermore, CA. Her name is Libby Mihalka CFA MBA. Ms. Mihalka is the founder of Altamont Capital Management. She has a blog: Financial Pragmatist in which she shares her thoughts. I read her May 9th article and it hit home. I like her analytical data-driven approach and the way she looks at data and historical trends. She brings up many great points in her writings. I went to her company site and devoured the rest too. The points I took with me are:

All in all, Libby thinks it is time to do some rebalancing and take some of the great returns from value stocks -- and especially small-cap value stocks -- off the table. Libby's chart of the week is reproduced above.

What does our model portfolio say

The good news: our model portfolio is still well ahead of the major indexes for the year. The bad news: it failed to outperform two of them (DJI and S&P500) in the past month. This is not surprising given the general overweight of small caps in our portfolio. The model is starting to score US large caps higher due to recent momentum. In particular, VTV is now at position 19, and IWD at 30. It may be time for some shift towards large cap from small and towards US stocks from small international stocks. Here are last Friday's rankings (top 30 ETFs):

	Using mmvr ranking method on 20070518
	1        3.3829 RPV     Rydex S&P 500 Pure Value
	2        3.0359 PRFE    PowerShares FTSE RAFI Energy
	3        3.0091 RYE     Rydex S&P EqWght Energy
	4        2.9445 DKA     WisdomTree Intl Energy
	5        2.8905 XLU     Utilities Select Sector SPDR
	6        2.7862 IDU     iShares Dow Jones US Utilities
	7        2.7280 EWK     iShares MSCI Belgium Index
	8        2.7032 PUI     PowerShares Dynamic Utilities
	9        2.6983 VPU     Vanguard Utilities VIPERs
	10       2.6791 EWG     iShares MSCI Germany Index
	11       2.6778 PRFU    PowerShares FTSE RAFI Utilities
	12       2.6732 RYU     Rydex S&P EqWght Utilities
	13       2.6385 VAW     Vanguard Materials VIPERs
	14       2.6238 FEZ     streetTRACKS Dow Jones Euro STOXX 50
	15       2.6135 PWV     PowerShares Dynamic Large Cap Value
	16       2.5840 UTH     Utilities HOLDRs
	17       2.5424 RFV     Rydex S&P Midcap 400 Pure Value
	18       2.5174 PXE     PowerShares Dynamic Energy Exploration
	19       2.4763 VTV     Vanguard Value VIPERs
	20       2.4706 DFE     WisdomTree Europe SmallCap Dividend
	21       2.4693 EZU     iShares MSCI EMU Index
	22       2.4564 PRFM    PowerShares FTSE RAFI Basic Materials
	23       2.4521 EWY     iShares MSCI South Korea Index
	24       2.4480 JKI     iShares Morningstar Mid Value Index
	25       2.4276 KIE     streetTRACKS KBW Insurance
	26       2.4132 XOP     SPDR Oil & Gas Exploration & Production
	27       2.3935 JKF     iShares Morningstar Large Value Index
	28       2.3852 DBN     WisdomTree Intl Basic Materials
	29       2.3530 PRF     PowerShares FTSE RAFI US 1000
	30       2.3278 IWD     iShares Russell 1000 Value Index

I must mention that I no longer recommend the Rydex "Pure Value" ETFs since I discovered they have abysmal liquidity resulting in high impact cost (bid/ask spreads, strange unexplained jumps when you want to buy them, etc.) so skip them. Energy, materials, and utilities had great runs but they still look good.

Is this an attempt to time the market?

Timing seems to be a dirty word. I've long convinced myself that I have absolutely no prediction power for short term moves -- daily and weekly moves look almost random to me -- but that the longer moves can be exploited by anticipating them (based on statistics, and history) and by giving up a part of the move, which means acting only after a new trend has been established, rather than trying to guess exact tops and bottoms.

This is timing in the same sense that rebalancing of a portfolio is timing. Tactical asset-allocation is done after a large move has been made, in order to restore a portfolio to its target allocation percentages.

Some people would say "it is time to rebalance". I take it a bit further saying "it is time to underweight small-caps". I feel certain enough about it to do it in all of my retirement accounts (where there's no tax impact), and to a lesser degree in taxable accounts too.

And that's the horoscope for today.

-- ariel


A special acknowledgment:
This is an opportunity to thank Lee Nelson who while studying finance at Stanford introduced me to so many of the concepts described above; Fama and French, Modern Portfolio theory, rebalancing and asset allocation, trying to read tea leaves in random charts, and much more.