In a previous article (read here) I proved that a simple market timing technique could bring an additional return and a lower risk to a dividend portfolio, taking into account lost dividends and transaction costs. But even with a proof, most income and dividend-growth investors are reluctant to sell and buy again a whole portfolio, especially when the capital or number of stocks is big.
For them, hedging is another solution. But hedging all the time has a cost, especially for dividend stocks. They often lag the S&P 500 index in bullish periods, and outperform it in downturns. Hedging them on the benchmark index will boost the return in bear markets, and scratch it in bull markets.
How to fix it? You can use a market timing signal to detect risky periods, but instead of selling stocks, you take a temporary short position in SPY or a long one in SH. This is less hassle than selling and buying again all positions, less transaction costs, and less frustration about lost dividends. An inconvenient: it needs leverage and margin.
I wish to illustrate the theory with a dynamic strategy I'm using on a real account. After a short description, I will compare the standard strategy and versions with market timing, hedging 100% of the time, and timed hedging. The aim is to measure the impact of protection tactics on a reference portfolio. The strategy itself cannot be detailed here.
I list the S&P 500 stocks whose last registered dividend yield was 3.5% or more, and whose short interest is below 15%. Then, I apply a quantitative ranking using ratios on valuation, growth and debt. This is a pure fundamental strategy with a weekly rebalancing. There is no technical analysis involved. It gives an average annual return above 20% with the 25 best stocks on 15-year simulations. The following example uses only the five best stocks.
Market timing indicator:
I will use the same for both market-timing and timed hedging implementations. It is defined by a bearish signal when the S&P500 current year EPS estimate falls below its own value 3 months ago, and a bullish signal when it rises above this value. This is an aggregate fundamental indicator, not a technical timing. Of course, you can find better market-timing indicators. The choice of a 3-month period is common sense: it is a reporting cycle of fundamental data.
Simulated performance since 1/1/1999:
+ market timing
Hedged all the time
+ timed hedging
The Sortino ratio is a Sharpe ratio taking into account only the "bad" (negative) volatility. I include rates of 0.1% for transaction costs and 2% as carry cost on the leveraged part. Dividends are reinvested.
With an average return above 30% and a Sortino ratio above 2, the timed hedging version looks impressive. Here is the equity curve:
Looking at individual holdings is a risk to miss the big picture. This is a weekly strategy based on quantitative fundamental analysis and Standard & Poors / Compustat data. Current holdings are of minor interest for the demonstration. Nevertheless, to make the case realistic, the next table shows them at the time I write this.
Oil, Gas & Consumable Fuels
Energy Equipment & Services
Seagate Technology Plc
Computers & Peripherals
As I wrote earlier, the strategy works also very well with up to 25 holdings. Adding diversification rules is possible, but out of scope here.
Other choices for hedging:
The portfolio is hedged and leveraged twice about 30% of the time. To reduce the carry cost, it is possible to use leveraged ETFs, long or short. My next article will show you which ones and quantify their natural decay (beta-slippage). You may want to click right now on "Follow" (on the left) if you don't want to miss it.
Conclusion: it is safe, but not safe enough
Implementing a good dividend strategy with timed hedging is probably the safest way to invest in stocks. But it is not safe to invest in a single rationale, even the best one. The usual disclaimer says: "past performance is not a guarantee for the future". First, because data are never enough to make a statistical model. Second, because even a perfect statistical model is no guarantee. Gamblers call this phenomenon "luck", scientists call it "non-ergodicity". The only way to be luckier is combining various strategies with an appropriate money management. Readers wanting more details can find links to my book and web site in my Seeking Alpha profile.