It'sa good question. Without knowing all the information and doing lots of analysis here are some things I would consider:

1) Lock away most of the original profit and keep the system going on small bankroll. This will give you time to observe the trend and protect the bank if it is going pear shaped. If it comes back on track you can re-add your money.

2) Do a month by month statistical analysis. Get the averages and stats for each month. Then look at recent months and see how far (if it all) they are below the worst month. It might turn out you've just had a run of bad months but could expect more big months coming in.

3) Using the data from 2, if you notice a distinct point in time where a change happened do some research on the events at that time. Maybe conditions changed (eg qualifications for a race type got stricter, or handicapping was relaxed in XYZ races) or maybe it was something like weather.

Plenty more things like that you can do. The main thing is do the research and see if you can explain the difference. Even if you have to throw the system away at least get confident in your decision first.