Individualized inferential expectations

[sbenthall]

AdaptiveExpectations should generate expectations of the risky asset based on individualized experiences of the price data based on when they are tapped by the Calvo Fairy.

This will require the agents to keep track of the days they have been tapped, and compute the risky_expectations based on a filtered history of the market price data.

@llorracc to provide a specification for this.

[sbenthall]

Maybe... FinanceModel.risky_expectations() takes 'attending_days' as argument.

[llorracc]

quick point: "Adaptive Expectations" is an existing terminology in economics which means something different from what we are doing here, so we'll need to have another name for this.

[sbenthall]

We have gone with the name 'inferential expectations'. any references to adaptive expectations has been obsolesced.

[llorracc]

There are two different things.

One is the fact that people do a hypothesis test -- that's the inferential expectations thing. The second is that they know past prices at only a limited (and random) set of past dates. I thought that it was this latter that you were proposing to call the "adaptive" part.

On Fri, Apr 28, 2023 at 2:03 PM Sebastian Benthall @.***> wrote:

We have gone with the name 'inferential expectations'. any references to adaptive expectations has been obsolesced.

— Reply to this email directly, view it on GitHub https://github.com/sbenthall/SHARKFin/issues/178#issuecomment-1527911057, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAKCK72CSYLCJSAK2EEZRHDXDQA7TANCNFSM6AAAAAASUX2HVY . You are receiving this because you were mentioned.Message ID: @.***>

--

• Chris Carroll

[sbenthall]

The second part I was calling "individualized". We now have individualized inferential expectations in SHARKFin as of #210 merged today. We are starting preliminary simulations this weekend. The main remaining issue is starting wealth. See #205