In 2012, Ben Jacobsen, Professor of Finance, Massey University of New Zealand and Cherry Yi Zhang, Professor at Nottingham University Business School, China, published a paper, Are Monthly Seasonals Reals?, analyzing 300 years of UK stock market data. Their conclusion was that a good case could be made for monthly seasonal anomalies in the stock market returns.

The professors analyzed 317-year index of monthly UK stock prices, UK 3-month T-bills sine 1900 and Bank of England lending rate since 1694. They found that the old market wisdom, “Sell In May and Go Away” works pretty well.

This phenomenon started because in olden times, the English upper class used to spend the winter in London and summer away from the stock market on their estates. The trend persists in the present times with the financial community continuing with the tradition of going on vacations during pleasant summer months.

The result is hat the winter returns show on an average 0.56% improvement from the summer month returns in UK. The summer returns have generally been lower than the risk free rate, suggesting a persistent negative risk premia over 300 years.

The professors conclude that traders with long-term horizon using trading strategies based on this market wisdom have remarkable odds of beating the market. Over a 5-year horizon, this strategy has beaten the market 80% of times and 90% of times over a 10-year horizon. Also, the average returns are three times higher than the average market returns.

We have used this strategy and reversed it – meaning we go long the broader market near the end of third quarter or beginning of fourth quarter and exit some time after the end of first quarter. For the rest of the time we stay invested in the treasury bonds. For entry and exit signals we use our proprietary algorithm, which uses a combination of oscillator indicators and support & resistance levels. The results are fantastic.

Using stock and bond index based ETFs, since 2000, our strategy has produced 133 trades with 81.2% success rate. The average return for an annual cycle – including both stock and bond trades – is 14.9%. The corresponding return for the market – considering S&P 500, Dow, NASDAQ and Russell 2000 – is 6.8%. The strategy was negative only two times out of 13 cycles. The broader market was negative four times.

The compounded return of the strategy for these years is 390%, giving 12.4% annualized return. The overall market indices produced 49% compounded and 3% annualized returns. However, strategy beat the broader market seven times on a cycle-by-cycle basis without taking the cumulative effect.

Some of the relatively bad years for the strategy were 1) 2002-2003 – the strategy returned 10.9% and the market 42%; 2) ’04-’05 – the strategy returned 0.5% and the market 7.4%; 3) ’08-’09 – the strategy returned 1.6% and the market 23.7%; 4) ’12-’13 – the strategy’s return was 12.2% and market’s 22.4%.

Some of the better years were 1) 2000-2001 – the strategy lost only -4.0% but the market lost -30%; 2) ’01-’02 – the strategy returned 5.3% but the market lost -16%; 3) ’09-’10 – the strategy gained 39.5% but the market only gained 12.7%; 4) ’11-’12 – the strategy gained 97.9% but the market lost -0.5%.

Breaking it down by stock and bond ETF trades, the stock market trades – including leveraged ETFs – have a success rate of 85.2% with average return of 11.3% for an average trade holding period of 200 calendar days. The compounded return is 237%. The bond related ETF trades have a win rate of 76% with 9% average return for a compounded return of 71% and average holding period of 141 calendar days.

So far the trades for 2013-2014 cycle are:

ETF	Entry Date	Entry	Exit Date	Exit	Ret
UWM	30-Sep-13	72.90	22-Apr-14	83.54	14.6%
IWM	30-Oct-13	109.83	22-Apr-14	114.62	4.4%
SPY	01-Nov-13	176.21	21-Apr-14	187.04	6.1%
QQQ	01-Nov-13	82.81	21-Apr-14	86.84	4.9%
SSO	01-Nov-13	92.68	21-Apr-14	105.16	13.5%
QLD	01-Nov-13	88.07	22-Apr-14	98.79	12.2%
DIA	07-Nov-13	155.85	21-Apr-14	164.03	5.2%
DDM	07-Nov-13	101.73	21-Apr-14	113.6	11.7%

The clock is ticking for the bond trades to kick in. As and when they are triggered we will send alerts.