Up, Down, Inside and Outside

Days Spent in Various Types of Price Action

I’m working on a new, relatively short-term trading strategy to add to my current mix of 9 strategies. Because of the use of daily data and my potential use of conditional orders with this particular new concept, I thought it would be useful to know how many outside days there were, since these days would typically produce potential losses with the proposed strategy. While I was at, I thought I’d run all four types of daily price action: Up, Down, Inside, and Outside days and pass it on to traders around the world.

What constitutes each type of daily price action?

Up Day:

If the today’s high is higher than yesterday’s high AND today’s low is equal to or higher than yesterday’s low, you have an Up day.

Down Day:

If the today’ high is equal to or less than yesterday’s high AND today’s low is lower than yesterday’s low, you have a Down day.

Inside Day:

If the today’s high is equal or less than yesterday’s high AND today’s low is higher or equal to yesterday’s low, you have an Inside day. Prices have stayed inside of the high and low of yesterday.

Outside:

If the today’s high is greater than yesterday’s high AND today’s low is lower than yesterday’s low, you have an Outside day. Prices have ventured outside of yesterday’s high and low.

Graphical Representation of Up, Down, Inside, and Outside Days:

So, How Many Days are Each Type of Price Action?

I took all the data on the SPY ETF (S&P 500 composite index Exchange Traded Fund) and dumped it into a spreadsheet, set up some conditional if statements and counted how many instance of each price action occurred over the period from January 1, 2000 to September 5, 2023, a total of 5,955 days.

The following summary is how they break down:

Up Days: 2,567 or 43.11%

Down Days: 2,128 or 35.73%

Inside Days: 668 or 11.22%

Outside Days: 592 or 9.94%

Total Days: 5955 or 100.00%

How I will use this information:

This helps me out in a couple of areas. In the simulator that I am using, the program is looking at daily highs and lows to see if a stop order is hit. Using daily data, the computer cannot know whether a conditional order was hit during the day after the primary order was executed or before it was executed. For a simple example, say I have an stop buy order to buy above yesterday’s high, then go live with a conditional order to sell stop the position below yesterday’s low. Prices may have gone lower first, moved higher and hit the stop buy and closed in the upper part of the range of prices for the day. The computer can’t tell whether the conditional price would have been hit. It has to assume it is executed and I was whipsawed in the simulation.

Knowing that only 9.94% of the simulation trades would see these Outside days, makes me comfortable. Some of these simulated 9.94% of the trades that the sim would call whipsaws, in many cases would not exist in the real world making the simulation a conservative look at what I might expect to see in the real world.

Second, knowing that over 78% of the days are trending up or down, I may look into using that to set up a trend-based strategy that assumes an up or down day for that ticker. I can have the strategy ignore Inside Days. There’s not many of them and they are days where the market doesn’t move much, so there’s not a lot of profit to be had. Outside days are not trending and will likely create losses for trending strategies, but since they are less than 10% of the days, I can deal with that mentally.

Third, traders that trade the trends may find some comfort knowing that over 78% of the days out there are trending up or down. That should give some comfort in multi-strategy approaches that rely on trending markets to produce profits and help traders enjoy the ride!