Halloween Indicator: Timing The Market Significantly Beats Time In The Market

Halloween Indicator: Timing The Market Significantly Beats Time In The Market

No Comment

by Kerry Balenthiran

Common wisdom says that there is no point trying to time the market, it can’t be done. Time in the market is more important than timing the market and is what matters from an investment growth standpoint. This view is backed up by countless academic studies and marketed by the fund management industry who want you to be invested all of the time.

To back this up, time in the market proponents often cite the example of what would happen if you missed the 10 best days of stock market performance – obviously you would under perform the market if you missed the best days. What they fail to say is that conversely if you had missed the 10 worst days, then you would outperform the market. In fact Mebane Faber of Mebane Faber Research has shown that had you missed the best and worst 10 days of the Dow Jones Industrial Average performance from 1984 to 1998, a period deliberately chosen as it was a very bullish period, then you would have still beaten a buy-and-hold approach.

The conclusion that I took away from this is that timing market is a better approach than passive investing as long as you can do it well. And therein lies the challenge; how do we go about market timing and ensure that we do it well? How do we consistently beat the buy-and-hold market return. I decided to look for a way to outperform the market using an objective market timing strategy.

Six Month Switching Strategy

In The UK Stock Market Almanac 2013, Stephen Eckett shows that a strategy of being invested in the market during the period 1 November to 30 April and moved to cash for the rest of the year, £1,000 invested in 1970 would be worth £46,529 by 2011, approximately triple the buy and hold return, excluding dividends. Whereas if we had done the opposite, been invested in the market between 1 May to 31 October and cash the rest of the time, the original £1,000 would only be worth £445.

Research by Ben Jacobsen and Cherry Y. Zhang of Masset University shows that seasonal market cycle has been evident in the UK markets going back hundreds of years and was first documented in the Financial Times in 1935. Eckett also shows that this strategy works with dividends reinvested. Starting with £1,000 in 1994, the six-month strategy would have increased in value to £4,884 by 2012. Over the same period, a portfolio tracking the total return of the FTSE 100 Index would have increased to £3,389.

US Presidential Cycle

In Trading Secrets, Simon Thompson shows how bear markets are most likely to bottom out in the first or second year of the US presidential cycle (presidential terms are fixed at four years). By combining the presidential cycle and the six month cycle, Thompson refers to research from Dr Marshall Nickles that shows a strategy of buying a US S&P 500 Index tracker on 1 October of the second year of the second presidential term and holding this position until 31 December of the year of the election. This strategy turned $1,000 into $72,701 over a 48-year period (1952 to 2000), with no down years. Unfortunately the next buy period, the 27-month period to 31 December 2008, resulted in a large draw down as it included the Lehman Brothers meltdown.

Dr Nickels chose 1952 as the starting point for his data because, in his opinion, the US government generally did not attempt to influence the economy in any significant way prior to 1952.

What about the period prior to 1952? Mebane Faber has shown that this presidential cycle strategy did not work from 1900 to 1950 and actually underperformed a buy and hold strategy during this 50-year period and from 1900 to 2010 overall!

Balenthiran Switching Strategy

All of this got me thinking about my own market timing model, the Balenthiran Cycle, described in my book The 17.6 Year Stock Market Cycle and wondering whether a simple objective mechanical trading model, one that removes emotions and subjective decision making, could outperform the market consistently. While I don’t know the exact dates of panics and crashes, I have demonstrated that they happen in regular intervals and how cyclical (short term) bull markets typically last 4.4 years and cyclical bear markets last 2.2 years, within a secular (long term) 17.6 year bull or bear market cycle.

Taking the presidential cycle as a starting point, but substituting the cyclical bear market phases of the Balenthiran Cycle as periods in which to be out of the market, I have devised a similar strategy.

For the sake of simplicity I will assume that the investor has a simple index tracker and ignore transactions costs, dividends, interest on cash deposits, taxes, etc., which is consistent with the other studies mentioned.

The strategy is for an investor to buy and hold during the entire secular 17.6 year bull market cycle, but to sell on 30 April after the secular top and move into cash. The investor would then re-enter the market during the first two cyclical bull market phases during the secular bear market, and move to cash during their subsequent bear market phases. The investor would then remain invested in the market during the remainder of the secular bear market and through the next 17.6 year bull market.

Got it? I thought not; let’s look at example of this.

Fig 1 Balenthiran Cycle

Figure 1: Secular Bull Market from 1982 to 2000 and Secular Bear Market from 2000 to 2018

Using the Balenthiran Cycle shown in Figure 1, covering the 1982 to 2000 bull market and subsequent bear market, our investor would:

  • Be invested in the markets throughout 1982 to 2000, buying on 31 October 1982 selling on 30 April 2000 (or the nearest trading day).
  • Stay in cash until 31 October 2002 when they would reinvest.
  • Sell on 30 April 2007 and move into cash.
  • Reinvest on 31 October 2009.
  • Sell on 30 April 2011 and move into cash.
  • Reinvest on 31 October 2013 and remain invested through to 2018 and on to 2035.

The results of the Balenthiran switching strategy are remarkable. Figure 2 shows that $1,000 invested in the Dow Jones Industrial Average (DJIA) in 1897 would have grown to $4,975,246, more than 19 times as much as buy-and-hold! In addition the strategy only delivered negative returns once, in 1940 (although the strategy missed out on gains in 1908 and 1938).

Fig 2 1897 to 2011

Figure 2: Balenthiran Switching Strategy versus Buy-and-Hold from 1897 to 2011

The red shaded dates are the time to sell, these are listed in my book going out to 2053, and the subsequent declines are shaded amber. The un-shaded dates are the times to buy-and-hold. For the more adventurous investor, I have shown what would happen had our investor shorted during the bear phases rather than just moved into cash. A long/short strategy turbo charges returns producing 121 times as much as buy-and-hold (Figure 3). To put this into context, in 1897 an ounce of gold cost approximately $20, so the starting investment in this example equates to roughly 50oz of gold, worth around $70,000 today.

Fig 3 Bal Switching Graph

Figure 3: Graph of Balenthiran Switching Strategy versus Buy-and-Hold from 1897 to 2011

The Balenthiran switching strategy compares well with both the six month switching and presidential cycle strategies. Dr Nickels showed that $1,000 invested in 1952 using his presidential cycle grew to $72,701. Over the same period the Balenthiran switching strategy returns $75,384 (Figure 4). Once costs are taken into account I think the Balenthiran switching strategy will be marginally even better as there is no switching during the secular 17.6 year bull cycles or the last five years of the secular bear market, resulting in lower transactional costs.

Fig 4 $1k vs US Nickel

Figure 4: Balenthiran Switching Strategy versus Buy-and-Hold from 1952 to 2011

In the Stock Trader’s Almanac 2012, Jeffery Hirsch shows that $10,000 invested in the DJIA in 1949 using the six month switching strategy returns $619,071 by 2010. Hirsch uses a slightly different strategy based on the presidential cycle and using the Moving Average Convergence Divergence (MACD) indicator to refine entries, and his presidential cycle strategy returns $1,647,124 by 2010. Over the same period the Balenthiran switching strategy returns $1,888,026 (Figure 5), thereby beating both strategies. Again, once costs are taken into account I think the Balenthiran switching strategy will perform even better, plus as the dates are fixed there is no need to watch over indicators wondering when the trade signal will arrive (no emotions).

Fig 5 $10k vs US Almanac

Figure 5: Balenthiran Switching Strategy versus Buy-and-Hold from 1949 to 2010

The most important difference between these strategies though is the fact that the Balenthiran switching strategy has worked since the start of the DJIA, not just the last 60 years. If a calendar-based phenomena is present then it should be timeless and we shouldn’t be selective about the dates we use.

Returning back to the FTSE, we can see that the Balenthiran switching strategy works in the UK too, returning just under double the buy-and-hold return of the FTSE 100 since inception (Figure 6).

Fig 6 FTSE

Figure 6: Balenthiran Switching Strategy versus FTSE 100 Buy-and-Hold from 1984 to 2013

These results clearly illustrate that the bear phases of the Balenthiran Cycle work in identifying, in advance, when to be out of the market in order to avoid the periodic crashes that occur. It is evident that it does not work every time, but it provides a valuable edge to tilt the odds in an investor’s favour.

I am not suggesting that you passively buy an index tracker and hold it until 2035, but if you do make sure it is a total return index tracker, i.e. it tracks the return of the index including dividends.

I developed the Balenthiran Cycle to allow me to confidently time my move into cash, which after all is a valid investment class, although it is often overlooked. I will be investing in growth stocks offering value using my fundamentally driven ValuableGrowth methodology, having been predominately in cash from May 2011. You should invest using whatever method you are comfortable with, whether that is buying a fund, selecting your own investments or buying a tracker.

But keep an eye on the calendar, because as we have seen, market timing works and the next buy date, 31 October 2013, will soon be upon us.

About the author

Kerry Balenthiran
Kerry Balenthiran studied mathematics at the University of Warwick and then worked as a Spacecraft Operations Engineer in the UK and at the European Space Agency. He qualified as a chartered accountant with Arthur Andersen and now works as a consultant within financial services. His mathematical background led to a fascination with the cyclical nature of stock market booms and busts.

No Comments


Leave your comment



Back to Top