Hello Ioannis,

Morningstar's roe data is updated quarterly, so you are getting the same value over multiple days because your factor's window length is not large enough to go back to previous quarters.

You will need to increase the window length of each of your factors according to the quarter you are trying to query. In general, you can use 65 as a good estimate for a quarter's length (in trading days). Attached you will find a copy of your notebook where I have updated Roe2 and Roe3 to query values for the previous two quarters.

I hope this helps.

Got it,
thanks Ernesto

Hey,
Still in my quest to properly construct a dataset inclusive of fundamentals and technical indicators.
Roughly speaking the following function obtains an indication of how much a security has move during the most current quarter (zero indicates today)

class Momentum(CustomFactor):
inputs = [USEquityPricing.close]
window_length = 252
def compute(self, today, assets, out, close):
out[:] = close[-65] / close[0]

If I want to find the move over the previous quarter, is this the correct way to go about it?
class Momentum(CustomFactor):
inputs = [USEquityPricing.close]
window_length = 252
def compute(self, today, assets, out, close):
out[:] = close[-130] / close[-65]

I feel like something is not right. Can someone please verify that I am doing the right thing?

Thanks

Hi Ioannis,

Actually, the most recent value would correspond to close[-1]. close[0] corresponds to the value at the start of the window, which would be 252 days ago.

Other than that, both factors look ok.