Rolling Test back online. Please re-run your tests

Announcements
1

Dear All, The Rolling Test tool announced here is back online. We hopefully addressed the crashing problem. But more importantly each sample was not producing the same result as manually running the same period. The bug mainly affected sims with custom universe. Sims with pre-built universes, like All Fundamentals, should not be affected that much. Please re-run your test to update the statistics.

Thank you for your patience and your feedback.

2

I don’t see the hedge selection control. Are hedges always disabled now?

Walter

3

The hedge controls only show up if the parent model has hedge enabled.

4

In that case, I should be seeing it and I’m not. Can anyone else confirm?

Walter

5

Previous test beta was using all the history data available on P123, current seems is obeying membership limits on historical data. I had a hope this was a shift to more historical data with free membership introduction. Any changes on historical data access pending?

6

I receive an error that the rolling server is offline.

7

Yes, sorry. It’s unstable at the moment under load. It does not affect anything else as it’s a separate machine.

8

Marco,

I know you guys are working on a ton… do you think you can add some summary stats for all the ‘rolling tests’ - average, median and STdev of 1 year returns?

Best,
Tom

9

It’s online again. We’ve addressed a stability issue that we hope will keep the server online.

10

I’m still having problems running a rolling sim on a model that uses a hedge. Now the rolling sim won’t start. It simply remains on the start screen.

Walter

11

Rolling tests with hedging is still an issue. Sorry about that

12

Rolling server is back online. We fixed another problem that caused the server to shut down. Rolling tests of hedged systems is still not right. Please de-activate the hedge when running rolling tests.

Very sorry for the inconvenience.

PS. can someone make a case for why you’d want to run rolling tests with hedging? I guess if it’s on 100% that could be justifiable, but anything else is just a test of market timing which is not very useful (# of samples is too small).

13

Marco - reason for me is what you said - constant hedging of a system vs. some index. I trade this on a lot of what I do, so it’s useful to test with rolling tool with hedge on.