May Seasonality: S&P 500

Sell in May and go away? Not exactly…

Click on the chart to enlarge.

From 1960 to 2010 Calculations start at the open of the first trading day of the month and end on the close of the last trading day.

May Statistics:

• Average Monthly Profit/Loss = 0.22%
• Winning Months= 56.86%
• Worst May = 1962 loss of -8.60%
• Best May = 1990 gain of +8.90%

Profit Distributions:

Equity Curve:

Since 1986 May has not really deserved the reputation it has gotten. It has not been a great month, but for the last 20+ years, it has been a fairly consistent (if small) winning month.

If historical patterns prevail, this May will be a month to consolidate the gains of the past few months.

Some new names were added for today. All the stocks from Thursday’s post are still valid.

All symbols are linked, and the charts will auto-update.

AMRN

SYMX

TZOO

The symbols are linked. The charts will update throughout the day.

BLTI

COOL

DYNT

INHX (PDS caught that large bounce earlier this month…)

VRUS

VHC

The market is extended here but one measure suggests an abnormally overbought market can keep going higher and higher.

I’ve written before, here and here, about an abnormally overbought S&P 500. I do not recommend shorting overbought markets as they have the tendency to stay overbought for longer than we might expect.

The Setup:

• Buy SPY (or $SPX) at the close when it closes above the upper Bollinger Band (built around a 50 period mean with the bands 2 standard deviations from the mean).
• Sell X days later.
• No commissions or slippage included.
• All SPY history used. $SPX history starts in 1960.

The Results:

While many traders see an abnormally overbought market and think it is time to short, this study shows that at the very least, it is better to just do nothing, or stay long.

While the edge is not much better than Buy and Hold (for the SPY), there doesn’t appear to be much value in expecting a major correction or even a big pullback. Perhaps if this setup was combined with other factors, such as low volume, we might predict when an abnormally overbought market might fail.

This setup, as shown with 50 years of data on $SPX, has consistently led to more market melt ups.

Some of these are giving a second opportunity to buy in…

Click on the charts to enlarge.

Should a large SPY gap-up be faded or bought?

On April 20th (or 3 days ago at the time of the writing of this article), SPY gapped-up over 1%. Not only was the gap-up large, but the low for the day was less than a dime beneath the opening price.

Gaps can be measured a couple of different ways, and for these tests, I’ve measured them from the previous close to the open and from the previous close to the low of the day. For example, SPY could have an opening gap (measured from the previous day’s close to the open) of greater than 1%, while the gap measured from the previous close to the day’s low would be less than the amount of the opening gap. Measuring the gap from the previous close to the low will be more restrictive and result in fewer trades.

The Rules:

Buy SPY at the close if

• The distance from the previous close to the open is greater than 1%
• The distance from the previous close to the low of the day is greater than 1%
• Sell X days later
• No commissions or slippage included
• All SPY history used

The Results:

Summary of Results:

The blue line measures the gap from the previous close to the low, while the  red line measures from the previous close to the open. Remember that there are many more gaps that are greater than 1% from the previous close to the open (179) compared to using the low (57).

The green line is a very simple measure of buying and holding SPY over X number of days, starting with the first day of SPY history. This measure is limited, to be sure (for example, altering the first day of data by a month or so will affect returns), but it does give a baseline from which to compare the gap trades.

Compared to the baseline, after a large gap up, SPY under-performs. However, the average trade low over the next 20 days (for both measures of gaps) is greater than the gap amount of 1%. This suggests that a gap-fill may occur intra-day, but on a closing basis, we are better off NOT betting on a gap fill over the next 20 trading days. Put more simply, large gap-ups are not easily filled.

After 13 trading days, buying after a large gap-up has tended to out-perform the baseline SPY performance.

Bottom Line: Fading large gaps with the hopes of profiting from a gap fill may not be the best way to play. Instead, it may be better not to fade, and instead buy weakness after the gap-up.

I would normally exclude a stock such as AMRN with a large gap-up. Usually large gaps mean a stock has been bought-out. In this case, AMRN received favorable data on a pending drug.