BSM tests

[alecloudenback]

Why is the test not passing?

[kasperrisager]

The build is complaining about LogNormal not existing. Is it just a namespace issue?

[alecloudenback]

The build is complaining about LogNormal not existing. Is it just a namespace issue?

I fixed the PR to capture the tests I was doing locally and added some comments on what I was trying to test

[kasperrisager]

I fixed the PR to capture the tests I was doing locally and added some comments on what I was trying to test

I can't seem to see that in this pr. Am I looking in the right place?

[alecloudenback]

I can't seem to see that in this pr. Am I looking in the right place?

My bad, forgot to push the commit.

[kasperrisager]

The p value of the test is 5e-128, so something is way off. I think it's the mean of the lognormal that needs to have m.q subtracted. Also, I don't remember the formula precisely, but sigma^2 are almost always divided by 2 when seen in the wild 🙂.

Also, the expected distribution is only lognormal in the limit of infinitely small timesteps, so with 60 timesteps, 10,000 replications might be enough to reveal that the distribution is a product of 60 normal distributions.

[codecov]

Codecov Report

Merging #12 (0d034ea) into main (44d8802) will increase coverage by 0.66%.
The diff coverage is 83.63%.

@@            Coverage Diff             @@
##             main      #12      +/-   ##
==========================================
+ Coverage   81.48%   82.14%   +0.66%     
==========================================
  Files           2        4       +2     
  Lines          27       56      +29     
==========================================
+ Hits           22       46      +24     
- Misses          5       10       +5     

Impacted Files	Coverage Δ
src/Yields.jl	0.00% <0.00%> (-100.00%)	⬇️
src/equity.jl	71.42% <71.42%> (ø)
src/interest.jl	80.00% <80.00%> (ø)
src/EconomicScenarioGenerators.jl	100.00% <100.00%> (ø)

📣 Codecov can now indicate which changes are the most critical in Pull Requests. Learn more