MSFT dividend data has gaps

[fashandge]

I downloaded dividend data using the rest API and use them in my application. While debugging the data using Claude Code, it finds the following issue on MSFT dividend data. Could you fix them?

MSFT pays quarterly (Feb, May, Aug, Nov), but Massive's data for 2017 has only two records instead of four — the May and November ex-dividend dates are
missing entirely. Their haf values bundle two quarters of adjustment into each event.

The consequence for adjust_daily_bars.py: for any price date between 2017-02-14 and 2017-08-14, the next ex-div in our data is 2017-08-15 (haf=0.908207).
Yahoo has 2017-05-17 as next, with a haf that's lower by one dividend factor (~0.6%). So our adjusted prices are ~0.57–0.60% too high for that entire
window. Same story between Aug and Nov 2017.

This is a source data gap in Massive — nothing in the pipeline is broken. Everything from 2018 onward is fine.

[fashandge]

I debugged further to find more tickers likely with quarterly dividends but missing dividends events for some quarters in the downloaded dividends data using rest API:

       ticker   ex_date        gap   combined_ratio  typical_combined_ratio   ratio
       -----------------------------------------------------------------
       TG       2020-12-21     186         0.703913    0.992918  0.7089
       SPQ      2024-12-23     181         0.862442    0.993475  0.8681
       CRDE     2022-08-15     182         0.880000    1.000000  0.8800
       CSR      2018-12-31     200         0.890625    0.987485  0.9019
       FTGC     2022-12-15     174         0.905180    0.992239  0.9123
       TNL      2018-06-14     202         0.902844    0.987850  0.9139
       ADX      2019-11-22     190         0.923513    0.996313  0.9269
       CWPS     2023-01-12     182         0.932581    1.000000  0.9326
       TFC      2019-08-13     182         0.923209    0.986799  0.9356
       SPUU     2017-06-20     182         0.934067    0.996133  0.9377
       FDLS     2024-12-16     174         0.941926    0.997680  0.9441
       LFVN     2023-09-07     191         0.941940    0.994506  0.9471
       FAAR     2022-12-15     174         0.943116    0.992174  0.9506
       RTX      2019-11-14     182         0.944148    0.992612  0.9512
       BRZU     2021-12-21     182         0.946297    0.993694  0.9523
       MYLD     2025-03-20     181         0.950500    0.993460  0.9568
       PEO      2019-11-22     190         0.952366    0.993365  0.9587
       BVERS    2022-09-21     182         0.959677    1.000000  0.9597
       PGJ      2024-06-24     189         0.957687    0.997041  0.9605
       BKNIY    2024-12-03     155         0.960770    1.000000  0.9608
       TTT      2024-12-23     180         0.956103    0.994886  0.9610
       DVYE     2024-12-17     189         0.948727    0.986719  0.9615
       ADVE     2024-12-18     175         0.957634    0.994404  0.9630
       FRO      2023-03-15     181         0.940749    0.975931  0.9639
       CMTOY    2023-04-06     182         0.965568    1.000000  0.9656
       VIASP    2024-03-28     182         0.966135    1.000000  0.9661
       CWPS     2022-04-13     181         0.966833    1.000000  0.9668
       HEQT     2022-12-27     183         0.965354    0.996650  0.9686
       NXTI     2024-12-23     181         0.968119    0.998125  0.9699

Explanation of columns:
gap: the number of days to the previous dividend event. If it's > 90 days, then it's missing dividend events for some quarters.
--- the following columns are for reference and debugging
ratio: calculated based on historical_adjustment_factor, today's factor / next event's factor
combined ratio: if there are multiple dividend events on the same day, we multiply those ratios