The Time Parameter: Why Time of Day Decides Your Breakout Results

I was standing in the pit at the Chicago Board of Trade when the noise stopped.

Not slowed down. Stopped. The screaming, the arm-waving, the whole circus went quiet, and I honestly thought trading had been halted. I turned to my guide, a veteran pit trader, and asked him what had happened.

“Lunch time,” he said.

I did not understand. In my head, America was the place where money never sleeps, where traders are machines who do not eat and do not blink in case they miss a move. He started laughing at me.

“Traders are just people.”

Then he left for his own lunch.

What did a lunch break teach me about trading?

That was the moment it hit me: there was no rational reason for me to sit in front of my charts for the entire trading session.

At the time I was still a discretionary day trader in Prague, and the regular session ran from 15:30 to 22:15 my local time. Nearly seven hours. I sat through all of it, every day, because I assumed that was what serious traders did.

But the big guys in the pit did not trade the whole session. So why was I? Why was I staring at a screen during their lunch, when nobody was trading?

What I saw in the pit confirmed exactly what my own trading had already been telling me. The most volatility, the most screaming and shouting, happened at the beginning and at the end of the session. In between was lunch, and then a boring, sleepy post-lunch stretch.

What did my own trade log actually show?

As soon as I got home from the U.S., I did something very simple. Three steps:

1. I opened the Excel spreadsheet of all my live trades. I kept very good records, down to the smallest detail.
2. I split every trade by hour of the session: first hour, second hour, and so on.
3. I looked at the distribution.

80% of my profits came from the first two hours of trading.

I could not believe I had overlooked it. I was furious with my own ignorance. All that work after the first two hours, another four and a half hours in front of the screen, for so little money.

I stopped trading after the first two hours immediately. Later I extended it slightly to the first 2.5 hours, which turned out to be my sweet spot.

Since that lunchtime on the CBOT floor, I have known one thing for certain: in trading, time of day matters. A lot.

When I moved to algorithmic trading, this was one of the very few things I carried over from my discretionary days untouched. That is why TIME is a mandatory component of the Mr. Breakouts formula.

How do you find your own best trading hours?

Two paths, depending on where you are.

If you already trade a market and know it well, do what I did. Pull your trade history and split it by hour. If you do not have enough history, use the VOLUME indicator and the ATR (Average True Range) to find which parts of the day are most volatile and offer the best opportunities.

Then build your breakout strategies with entries restricted to those hours. Even if the other hours are profitable too, limiting yourself to the best part of the session can meaningfully increase your Average Trade, cut down false breakouts, and improve your NetProfit/Drawdown ratio. Those three things are exactly what you should be optimising for.

If you are new, split the session into two to four equal segments. Build a few breakout strategies for each segment, then build the same strategies for the full session and compare. Is one part of the day giving you better trades, a higher Average Trade, a better NetProfit/Drawdown ratio? Trade that segment only, and keep measuring.

What does a time filter look like in code?

Here is how I test it. This is a simple long-only breakout on the e-mini Dow (YM), 20-minute bars, with the hour of entry as a switchable input:

Input:
    Fract(2.8),
    Filter_Period(45),
    Test_Hour(0); //0-6
vars:
    POI(0), SPACE(0), myBreakoutLong(0), FILTER(false), TimeFilter(false);

//Point Of Initiation
POI = CloseD(1);
//For the calculation of the space
SPACE = fract * AvgTrueRange(40);
//The BreakoutLevel (POI + space)
myBreakoutLong = POI + SPACE;
//Filter
FILTER = ADX(25) < Filter_Period;

If Test_Hour = 0 then TimeFilter = (time > 830 and time <= 930);
If Test_Hour = 1 then TimeFilter = (time > 930 and time <= 1030);
If Test_Hour = 2 then TimeFilter = (time > 1030 and time <= 1130);
If Test_Hour = 3 then TimeFilter = (time > 1130 and time <= 1230);
If Test_Hour = 4 then TimeFilter = (time > 1230 and time <= 1330);
If Test_Hour = 5 then TimeFilter = (time > 1330 and time <= 1430);
If Test_Hour = 6 then TimeFilter = (time > 1430 and time <= 1515);

//Entry Condition
If marketposition = 0 and FILTER and TimeFilter then buy next bar myBreakoutLong stop;
//Exit at the end of the day
if marketposition <> 0 then setexitonclose;
//Set stop-loss 1500 USD
Setstoploss(1500);

The last segment is 45 minutes rather than 60 because the market closes at 15:15.

Then I run a quick optimisation across Test_Hour 0 to 6.

What did the optimisation reveal?

Two things worth your attention.

First, restricting the strategy to the last 45 minutes of the session (Test_Hour = 6) made more money than the original strategy with no time restriction at all. This is not unusual. A time-restricted strategy beating the unrestricted version happens often.

Second, the lunch period proved my point exactly. Trading between 11:30 and 12:30 (Test_Hour = 3) lost money: USD$2,610.

One note before you panic at the arithmetic. The six time-restricted results do not add up to the Net Profit of the unrestricted version, because with time restrictions the trades are taken differently. A position can only be opened if none is already open, so some time-restricted trades simply would not have existed in the original. If that sounds confusing now, do not worry about it. Test it yourself and it becomes obvious fast.

That is the TIME component. It will often improve your results.

Not always. But often enough that skipping it is careless.