Stock Market Anomalies: The Time-of-Day Effect

In earlier research, we have considered several anomalies in the US stock market, such as:

• FOMC "drift" effect;
• the Tax Day anomaly;
• the Monday effect;
• the turn-of-the-month effect;
• the insider trading anomaly;
• the "Sell in May and go away" effect.

Today we will study the time-of-day effect.

Lawrence Harris, a professor of finance and business economics at the University of Southern California, was one of the first who studied this effect.

In his “Study of a transaction data study of weekly and intradaily patterns in stock returns”, published in May 1986 in the Journal of Financial Economics, he examined stocks of companies traded on the NYSE from 1981 to 1983 and came to the following conclusions: “For all firms, significant differences in weekday intraday returns accumulate during the first 45 minutes after the market opens. Prices fall on Monday mornings, while they rise on other weekdays. Otherwise, the picture of intraday return is the same on all weekdays. Most notable is the prices' growth on the last transaction of the day.”

US Stocks:

Strategy

1. Open long positions on the Open of a day.

Close positions on:

• on Close of the 1st 15-minute candlestick (in 15 minutes);
• on Close of the 2nd 15-minute candlestick (in 30 minutes);
• on Close of the 3rd 15-minute candlestick (in 45 minutes);
• on Close of the 4th 15-minute candlestick (in 60 minutes);
• on Close of the 24th 15-minute candlestick (in 6 hours - at the end of the trading day for stocks);
• on Close of the 96th 15-minute candle (in 24 hours - at the end of the trading day for indices).

2. Open long positions:

• at the opening of a 15-minute candlestick, 15 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 30 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 45 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 60 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 75 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 90 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 105 minutes before the end of the current trading day;
• at the opening of a 15-minute candlestick, 120 minutes before the end of the current trading day.

Close positions on the Close of the current trading day.

We will evaluate the trading strategy according to the following criteria:

• The average return reflects the relative change in quotations of financial instruments in percentage. A positive value of the average return indicates the profitability of the strategy, a negative one indicates a loss.

The average return (R) of a financial instrument is given by the formula:

where:

n - the number of transactions;

P (%) – the percentage of change in the quotation of the financial instrument at the time of fixing the position, is calculated as follows:

for buy positions

P (%) = (position closing price - position opening price) / position opening price * 100%

for sell positions

P (%) = (position opening price - position closing price) / position opening price * 100%

• The total return (TR) is the sum of the returns from all trades. The greater the value of the total return, the greater the profit brought by the signal during its testing period.

• Max drawdown (MaxDD) is the maximum loss in percentage terms from fixing losing trades for the entire testing period. The smaller the value of the maximum drawdown, the better the trading signal works.

where:

n - the number of trades;

R - rate of return;

TR_n - total return of n trades;

DD_n – drawdown at the time of closing the nth trade;

MaxDD – max drawdown.

Analysis of the obtained results 

Let's look at the results of testing the first strategy:

So, the hypothesis about the growth of US stocks in the first 45 minutes of trading from Tuesday to Thursday has been confirmed by the results of testing, the growth on Fridays and the decline on Mondays has not been confirmed.

It should be noted that on average, the growth of stocks during these minutes on Mondays, Wednesdays and Thursdays is about 50% of the total growth of these days.

The growth of indices in the first minutes of trading is less significant compared to stocks.

Let's estimate the total rate of return and maximum drawdown of the strategy of buying at the opening of the day and closing the transaction after 45 minutes:

Apple, Intel and Home Depot have shown the best results on Monday; Visa and Apple - on Tuesday; IBM and Procter & Gamble - on Wednesday; American Express, Home Depot - on Thursday; Visa and Procter & Gamble - on Friday.

Let's look at the results of testing the second strategy:

In this case, the hypothesis about the average daily growth of US stocks has been confirmed for entering the market 1.5 - 2 hours before the end of trading on Mondays and Fridays, about the growth of indices with the same entry-exit parameters has been confirmed, but on Tuesdays.

Let's estimate the total rate of return and maximum drawdown of the market entry strategy 2 hours before the end of trading:

Here, the leader of Monday is UnitedHealth, of Tuesday - Walt Disney, of Wednesday - Visa, of Friday - American Express.

Detailed results are shown in the Appendix:

XLSX (0.10 MB)Time of Day effect.xlsx