Databases Reference
In-Depth Information
The consequence of this learning over time is that the existing signals
are very weak. Things that were obvious (in hindsight) with the naked
eye in the 1970s are no longer available, because they're all understood
and pre-anticipated by the market participants (although new ones
might pop into existence).
The bottom line is that, nowadays, we are happy with a 3% correlation
for models that have a horizon of 1 day (a “horizon” for your model is
how long you expect your prediction to be good). This means not
much signal, and lots of noise! Even so, you can still make money if
you have such an edge and if your trading costs are sufficiently small.
In particular, lots of the machine learning “metrics of success” for
models, such as measurements of precision or accuracy, are not very
relevant in this context.
So instead of measuring accuracy, we generally draw a picture to assess
models as shown in
Figure 6-13
, namely of the (cumulative)
PnL
of
the model. PnL stands for Profit and Loss and is the day-over-day
change (difference, not ratio), or today's value minus yesterday's value.
Figure 6-13. A graph of the cumulative PnLs of two theoretical models
