Tick volume spikes: confirmation or just noise?

[Fairman]

Tick volume spikes: confirmation or just noise?

A short test most traders never run: take a month of breakout trades, split them into those where tick volume spiked at the break and those where it did not, and compare the results.

Plenty of people expect the spike group to do much better. Often it does, but not by as much as the folklore says. The reason is that a spike in ticks happens for lots of reasons, including a spread widening, a platform lag, or a series of tiny orders that move nothing.

Where tick volume tends to help is in ruling out. A breakout with no visible increase in activity at all is usually a bad bet, because there is little participation behind it. A breakout with a surge is at least a candidate.

So the signal is more of a filter than a trigger. It removes the worst setups and leaves a pool that still needs other criteria.

The honest answer comes from the numbers, not the reputation. A simple spreadsheet will show whether tick volume adds anything to the system in use.

A quick method for the test: tag each breakout with the ratio of the tick count in the break bar to the average of the previous twenty bars. Then group by ratio buckets, such as under 1, 1 to 2, and over 2. If the results are flat across buckets, volume adds nothing. If they climb with the ratio, there is a useful filter hiding in the data.

There is one more caveat. Tick volume is sensitive to the broker's feed settings, so a threshold found on one broker may not transfer to another. Re-testing after any change of broker is the safe policy.

Does filtering by tick volume actually improve the stats?

[FXS London Scalper]

A short test most traders never run: take a month of breakout trades, split them into those where tick volume spiked at the break and those where it did not, and compare the results.

Splitting a month of breakout trades by whether tick volume spiked at the break is a simple, telling test. Many find little difference, which suggests the filter adds complexity without edge.