My notion of a risk-off safe haven investment is presented here as a tactical fund of funds (TFF). A safe haven can be broadly defined as an asset that protects investors’ wealth against financial turmoil. it is commonly used to label very different assets, ranging from cash and Treasury Bills, to various risky assets like gold and commodities which can function as safe havens if they have negative correlations with the other assets in an investor’s portfolio during crises. Unfortunately that is often difficult to predict in that correlations often go to one during a crisis....depends on ability to time market since most safe havens follow stocks into the crapper.
Flying to safety is just another way of saying that investors are timing the markets — something has happened, and they are trying to get out of the market before it gets worse, with the idea of getting back in after the danger is gone.
12/31/17 by CS: When we start talking about defensive barbell strategies, I feel right at home. :-)
I have developed simple defensive bond strategies that can be used as the out-of-market (OOM) asset in PV. So rather than just using a single asset, we can use a strategy for the OOM asset. One modified defensive bond strategy I have developed uses OOSAX, VWAHX, and VFIIX and trades monthly. Here is the PV link that tests back to 2000 (limited by OOSAX; can be extended to 1990 if BFRAX is substituted for OOSAX): Test Market Timing Models. You will need to save this result as a benchmark in PV under your user ID, and give it a ticker name, e.g. OOM2000.
Next, use the volatility targeting tool in PV with XIV as the only asset (100% allocation). Set "use downside volatility only" = yes, set the volatility target to 3.7%, and set the volatility timing period to one month. The OOM asset is set equal to OOM2000. Here is the PV link: Test Market Timing Models. It won't run unless you have a benchmark file named OOM2000. See the previous paragraph to see how to create a benchmark file in PV. If you don't know how to do it, please send me a SA message and I would be glad to walk you through it.
This strategy trades monthly, giving more responsiveness. I obtain essentially the same results posted by varan, i.e. CAGR = 13.6%, SD = 7.9%, MaxDD = -5.1%, and Sharpe = 1.6.
But the volatility target strategy on XIV can be substantially improved if we use a two month volatility timing period rather than a one-month period. This makes sense since the volatility of XIV can change dramatically from one month to the next month. But when I insert two months in the place of one month for the timing period, the start date of PV moves to 2012 instead of 2011. So although I can show even better results if I use a two month timing period, the backtesting only goes back to 2012. Here is the PV link: Test Market Timing Models.
The volatility target can be increased to 5.1% when a two month timing period is used (still trades monthly) in order to match S&P 500 CAGR results. But now the volatility targeting strategy with the modified defensive bond strategy as OOM asset shows significant improvement in MaxDD (-4.1% vs. S&P 500's -8.4%) from 2012 - present.
Sales material by SJ: A Bear Market Strategy selects safe harbor investments during a market crash. When a market direction indicator (such as StormGuard or the Death Cross) signals conditions have become bearish, a Bear Market Strategy automatically takes charge and selects a trusted safe harbor investment from a list of candidates, such as cash, money market funds, bond funds, gold bullion, and US treasuries. Although numerous market sentiment indicators have been developed over the years to help determine when to flee to safety, none of them come close to the performance provided by StormGuard-Armor. While the theoretical basis for each of them sounds promising, many are only poorly correlated with future market performance. Comparison of some of the better ones is detailed HERE. The significant value Bear Market Strategies further add to StormGuard-Armor is illustrated by the remarkable difference between the yellow equity curves in the Strategy charts below. The reason Bear Market Strategies matter so much is illustrated by (A) the yellow equity curves, (B) the compound annual growth rates (OTCPK:CAGR), and (C) the Sharpe Ratios (risk-adjusted return) in each of the above charts. All three investment Strategies rely on SectorSurfer's True Sector Rotation algorithm to determine which one of the eight candidate SPDR sector ETFs to own each month during bull markets. The difference in Strategy performance comes from the addition of StormGuard-Armor and an integrated Bear Market Strategy. The Strategy on the left has no market crash protection; the Strategy in the center additionally utilizes StormGuard-Armor to determine when to exit the market to the safety of cash; and the right-most Strategy additionally utilizes StormGuard-Armor in combination with Bear Market Strategy BMS-4 to select the best trending ETF during bear markets from among safe harbor candidates UST, TLT, BND, MBG, MUB, CORP, SH, and GLD. Wall Street often refers to this as a risk-on / risk-off investment approach, where risk-on means a move to riskier investments with potentially higher yields during bull markets, and risk-off means a move to safer investments with typically lower yields during bear markets. In order to reasonably model the performance of a BMS (Bear Market Strategy), its candidate investments must have performance data that spans at least one major market crash. Although most ETFs that might provide safe harbor during a bear market were started only recently, most are based on indexes with much longer histories that can be used to artificially extend the ETF's data for purposes of improved strategy modeling.
from WSJ mark hulbert [link] To gauge asset classes’ likely performance during the next major stock-market decline, I analyzed all U.S. bear stock markets since the late 1920s. I relied on a list compiled by Ned Davis Research of Venice, Fla., according to which there have been 28 bear markets in that time. The asset class that has performed most consistently during bear markets is intermediate-term Treasury bonds. According to Morningstar data, intermediate-term Treasurys gained as a group during all but one of the 28 bear markets since 1929, and in the one bear market where they didn’t they still lost less than 1%. It is worth noting that interest rates rose in half of the bear markets since 1929, and this still didn’t prevent intermediate-term bonds from consistently producing a profit. During U.S. stock bear markets of the past three decades, for example, the S&P GSCI Index, a commodity benchmark, has fallen just as often as it has risen. The same overall conclusion applies to gold. Though it gained ground during the latest bear market for stocks, it also has fallen as often as it has risen during bear markets of the past three decades.
Start Date End Date Length in Months U.S. Stock Market Returns Intermediate-Term Treasuries Returns
1929-9 1929-11 3 -33.08% 3.37%
1930-4 1932-6 27 -79.56% 6.03%
1932-9 1933-2 6 -29.82% 2.06%
1937-3 1938-3 13 -50.04% 3.07%
1939-10 1940-5 8 -25.72% 3.93%
1941-9 1942-4 8 -22.38% 0.50%
1946-6 1946-1 16 -21.76% 0.33%
1962-1 1962-6 6 -22.28% 2.46%
1968-12 1970-6 19 -29.25% 2.63%
1973-1 1974-9 21 -42.63% 4.90%
1987-9 1987-11 3 -29.53% 2.38%
2000-9 2002-9 25 -44.73% 27.48%
my safe havens draft figure from 2010
Selecting 2 of the 5 candidates using a DM model worked very nicely for the 2008 bear market, returning 30% that year (PV link).
Long-term US Treasuries have been the best performing asset class during recent bear markets because they have been more negatively correlated with US Stocks than anything else for nearly two decades. However, the two have not always been negatively correlated, as is easily appreciated from the Rolling 5-year correlation between US stocks and 5-year Treasury chart (right). This means that they could become positively correlated again at some point in the future. Thus, the consideration of only long-term treasuries for a bear market strategy, based on their 20-year prior performance, would clearly constitute hindsight selection bias, resulting in substantially increased portfolio risk during the next market crash.
Only it rarely works that way. As the list above shows, predicting what will actually spark the next bear market is not an easy task. And don’t forget that to be successful in this endeavor, you need to not only avoid the danger, but get back into the market afterward.
You’re better off moving away from the all-or-nothing mind-set of fight or flight. Taking flight hasn’t served investors in the long run, but sticking to your long-term plan shouldn’t be a fight either. Rather, the right response is usually to be patient and realize that the apparent threat often is not as dangerous as first perceived.
Still, the flight-to-safety instinct is a strong one, and we eventually will see another bear market, so let’s discuss (and dispel) common funds investors turn to for safety. There are plenty of naysayers and doubters, but until proven otherwise, U.S. Treasuries remain the flight-to-safety asset of choice. You’ll get the most bang for your buck in terms of protection from falling stock markets in long maturity Treasuries.
table listing stock bears show monthly drawdowns of S&P and alternatives that bear market period
US Stock Market Bear Period
||S&P 500 Drawdown||CS's MDBS Conservative||
JL's ROFTS Option_1
JL's ROFTS Option_2
||0.00%, n.a.||0.00%, n.a.|
|1994 Bond Selloff
|1997 Asian Crisis
|| 0.00%, n.a.
|1998 LTCM, Russia
||0.00%, n.a.||0.00%, n.a.|
|2000-2 Dot Com
|2010 Greek Debt
|2018 Coiled VIX