This is the first two part article where I want to discuss the best investment strategies for next year. In the first part, I will cover my key market predictions for next year. In the second part of this article, I will discuss specific investment ideas: EZCorp - a struggling pawn shop and pay-day lender, Biolase - a dental laser company (special thanks to my wife Dr. Malkin) for helping understand its products, Perion - a small Internet software company about to get much bigger, and two emerging market small cap ETFs: RSXJ and SCIF.
As Yogi Bera supposedly once said: "it's hard to predict, especially, the future." I'd like to limit my market predictions to just two, which are backed by data, as opposed to making provocative but wild guesses what may happen in 2014. I predict that the US market will stay flat, even in the face of the improving economy, and a large number of this year winners will become losers and many losers will become 2014 big winners. Or putting it into investment strategy terms: buy what lagged this year and look for good investments outside of the US market. As investor in small-cap space, I find very few bargains (part 2) among small companies which rallied over 30% this year (IWM).
Prediction 1: US economy will perform in line next year with sub-3% growth but the US markets will be flat or even down
Bob Farrell: "When all the experts and forecasts agree - something else is going to happen"
Predictions of top strategies for next year - all smiles! (Source: Barrons)
Last week, Barron's published a summary of top 10 Wall Street strategies with their predictions for next year. There is a rather striking uniformity of opinion of what the market will do without a single dissenter. Having read Barron's for many years, most strategies are usually very bullish and display a clear herd mentality, especially at the top.
On average, the cheerful forecasters agree that S&P 500 (SPY) will go up another 10% (as of a week ago). The 10 year Treasury yield (TLT) will reach about 3.5% and the US economy will grow about 2.7%. The strategist with the biggest smile, Stephen Auth of Federated Investors, is especially bullish thinking S&P will hit 2,100. Of course, everything is possible, especially if we blow another asset bubble or what analysts euphemistically call "P/E ratio expansion", but let's look realistically if S&P can actually grow earnings from $109 average to $118 as expected on average (8.25% nominal growth).
Let's assume that they are right about the 2.7% growth of US economy, we can even make it 3%, since S&P 500 index companies export about 30% of its output (IMF estimates World Economy will grow 3.6%). We assume the inflation will reach the Fed preferred target of 2%. To reach the predicted 8.25% nominal growth earnings growth, the profit margins of S&P 500 companies will then need to improve another 3.3%-3.6%.
Historical profit margins (Source: Yardeni Research)
(click to enlarge)
As we can see from Yardeni Research paper, the margins are already record high. The great after-tax margins post-2008 crisis have been driven by three components: cheaper and more productive labor (workers had to do more work for less money), lack of capital investments (many offices are still running dinosaur Windows XP), and, finally, carry-over tax losses from 2008 crisis as well as various temporary tax breaks (mostly used up by now). None of these components are sustainable long-term so the profit margins are likely to slowly decline to its historical average of about 10%.
Let's assume that the margins will fall from 18% to a more reasonable, but still very high 16% in 2014. This 2% fall in margins would create a profit drag resulting in only 3% nominal profit growth (approximated as 3% real growth + 2% inflation - 2% margin drag). This 16% would be a mid-point between the peak margins in 2007, just before the crisis, and the current margin, as I expect the margins to gradually decline as the wage inflation and other costs slowly rise.
Therefore, I expect S&P 500 to earn at most $112 next year, not $118 analysts projected. If we assume more typical Treasury yields and apply a historical P/E ratio of 14.5, we get S&P 500 fair value only at 1,624, almost two hundred points below 12/19 close. If we assume the flat market prices and the nominal earnings growth of about 5%, it will take about three years for earnings to reach S&P 500 fair value.
Historical P/E Ratios (Source: WSJ)
Even if we look at the rosy projections of the market strategists, this graph shows a stark diversion between what companies earn and what investors pay for them. One can clearly see that the earnings growth has been flattening out since 2011 while the market kept accelerating.
US Market is ahead of US earnings estimates (Source: FactSet)
(click to enlarge)
The optimists will point out to the low Treasury rates and zero short-term rates driving investors into stocks. There is certainly some truth to this, but let's estimate the fair value of S&P 500 taking 10 year Treasury rate into account. We can calculate "implied equity risk premium" - the extra return investors expect to get from equities over a "safe" investment in 10 year Treasury where the only real risk is long-term inflation. I highly recommend Aswath Damodaran paper on explaining all gory details how it can be derived. He currently estimates that ERP is at 5.32%, below its long-term historical average of 5.88%, but above last 40 year's 3.91%. If we assume that the 10-year rates will move to 3.5% from current 2.9%, the risk premium will shrink even further, making an argument that stocks are cheap compared to bonds rather weak.
I realize that the negative market returns are relatively rare in absence of a recession due to investor psychology of not selling when things are looking up (there has been a single negative return year in 1994 since the 70's when not leading up to a recession). Therefore, I expect S&P 500 to stay flat for next two years waiting for earnings to slowly catch-up with plenty of volatility along the way. In absence of a strong market rally, which "lifts all boats", active investors are likely to do well while passive investment will produce scant returns.
My conclusion: S&P 500 companies will earn $112 next year and the index will finish the year at about 1,800, followed by two other years of flat or minimal returns. One should not put money in a passive investment such as SPY ETF.
Predictions 2: Many big losers will become big winners and many big winners will become big losers
Year-end tax-loss harvesting is a well-known phenomenon. Investors try to unload losers to offset gains from winning positions, especially in strong market years.
I believe that this year the effect of tax-selling will be especially significant resulting in an anomaly in January of investors selling winners and buying losers. Many investors are sitting on big gains for last four years after all major indexes this year reached record highs. Selling this year would expose them to taxes in April, selling after January 2 would delay the tax man by one year. It's important to note that the real tax selling starts at the end of November, as many investors want to buy the losers back next year but must wait 30 days to avoid a "wash sale" rule when the tax loss is disallowed if a stock bought sooner than 30 days.
To try to test this theory, let's compare the best performing ETFs with the worst performing ETFs this year in periods of last 30 days and a full year so far. I picked Internet (FDN) and Consumer discretionary (XLF) as the "winners" and Gold miners (GDX) and Coal industry (KOL) as the "losers" and SPY ETF as the "market". I also picked the best-performing Dow component HP (HPQ) and the worst-performing Dow component IBM (IBM), especially a good fit since they are direct competitors facing the same sector head and tailwinds. I constructed a hypothetical adjusted "pair trade" where I bought long the winner and shorted the loser and hedged with SPY ETF in the same dollar amount to remove the market risk. I compared how this trade would have performed year to date vs. last 30 days where I expected "tax selling" and then adjusted it for the same number of days. Even disregarding compounding, which would have made the result even more convincing, every pair trade outperformed in 30-day period its full-year-to-date equivalent.
For example, HPQ/IBM adjusted result is calculated as (17.8% + 3.2% - 1.3%)/[(365-13)/30]
(click to enlarge)
To be fair, this is somewhat anecdotal evidence, as I couldn't come up with precise methodology to run a statistical test but the evidence of winners outperforming losers toward the year end is certainly there. Perhaps, gentlemen from Bespoke Investments Group can backtest this idea in full.
My conclusion: the end of the year tax-selling effect is especially pronounced this year. It's likely that early in the year investors will sell the winners to lock-in overdue gains and buy the losers. You are well-advised to load up on losing stocks on December 31 and hedge it with puts on some high-flyers such as NFLX, AMZN, or PCLN.