We are often asked why our Transaction Cost Analysis (TCA) is primarily based in ticks (minimum price increments) as opposed to basis points (bps). For those experienced in equities, they’ll be familiar with bps as the conventional unit for equities TCA. When we were initially building QB, we first looked for an external TCA provider in the futures space but could not find one, promptly realizing that we had to build one to showcase our performance and assist our clients. With this fresh start, we decided to take an approach that we thought most logical and robust through all market conditions, as well as across a diverse universe of futures instruments. This approach allowed us to start from scratch and not simply use bps just because that was what was done in equities. In both our execution algorithms and TCA, we wanted to do what was right and most appropriate for the markets we were focused on. This approach led us to settle on ticks as the preferred unit for slippage measurement.

The central challenge with any measurement is precision. This requires one to measure accurately, to use the most precise instruments and remove any avoidable noise. At QB we take great care in ensuring we use the best instruments. Our timestamps and market data are captured as precisely as possible. For many this is where measurement begins and ends. However, it is also important to consider what the objective of that measurement is. In the case of assessing execution performance it’s not simply how precisely one can measure the slippage of any given trade, but also how this performance varies through time.

This is the biggest weakness of using basis points as a unit of measure. Looking at the equation below, a key component for calculating basis points is the Benchmark Price. Whilst this is certainly relevant for valuing one’s portfolio it has little bearing on the underlying market microstructure that drives slippage. By introducing this in our measurement one is adding noise; which at times can be very significant:

Slippage in bps = ( ( Execution Price – Benchmark Price ) / Benchmark Price ) *10,000

Slippage in ticks = ( ( Execution Price – Benchmark Price ) / Minimum Price Increment )

Example 1 – Crude Oil: The price almost halved in under 3 months. So for this reason, the basis points measure is misleading. Is the trader whose slippage was 3.5bps in January really 150% worse than the one who achieved 2bps in October?

Example 2 – DAX: A similar story. Did execution performance improve between January and June or did the denominator just get bigger?

Ticks are ticks, and are always the same value unless there is a tick size change instigated by the exchange (a relatively rare event).

The only plausible explanation for not using ticks to measure slippage is when the portfolio manager is interested in knowing the trading cost in dollars so as to compute the net return of the portfolio. This again is in dollars and not necessarily basis points. Furthermore, unlike U.S. stocks, futures have varying tick sizes and hence would need to be taken into account for the computation of execution cost. Price may even be the simplest and purest measure of all – literally the absolute price differential. However, one drawback is it makes comparing slippage across different instruments harder to comprehend compared with ticks.

In summary, why ticks make sense for determining performance:

Insulated from large market moves over time which can distort the bps measure.
Despite wide ranging tick values across various futures instruments, slippage in ticks gives a normalized measure to better understand and compare across different markets. For example, bps slippage in a large tick market will look bigger than a small tick market, but execution performance across both may actually be similar.
For rates markets, basis points slippage measurements will be very small and may also cause confusion with yield which is the usual basis point measure associated with these markets.
Slippage measurements are inherently noisy statistics as it is. Dividing by a random price simply introduces more noise without any benefit in understanding execution performance over time.

Quantitative Brokers
New York
July 8, 2019

Best Execution: Unit of Measurement