MYTH #5
If 99.x% of my timestamps fall within RTS 25,
the odds that regulators will discover a timestamp that
is out of compliance is very small.
The myth in this statement is its premise: that regulators will only investigate your RTS 25 compliance if they come across an errant timestamp. That may happen on occasion, such as if a trading firm’s timestamps indicate a different sequence of events than an exchange’s timestamps. More often, however, regulators will assess compliance with RTS 25 by reviewing the processes by which you ensure accuracy. Whenever they ask you to supply records of reportable events to support an investigation, they will also ask for your evidence that the timestamps on those events comply with RTS 25. They will ask themselves: is the firm performing “relevant and proportionate monitoring” and “relevant and proportionate testing”? If so, what do the monitoring results and test results show?
As part of their assessment, regulators will review your application-level error test results. And, assuming they approve your test methodology, they will decide whether you were right to draw the line on compliance where you did.
It’s worth noting that this means regulators will be unable to escape the question of tolerances themselves. As soon as they face a situation in which the firm’s testing records show a platform whose max application-level error makes it possible for max total timestamp error to exceed RTS 25, they will be forced to take a position.
The final mystery of interpretation involves some philosophy and some arithmetic….
<Next: MYSTERY #2 - Will regulators accept a skew-adjusted approach to application-timestamp error?>