For an easy understanding of the topic in this thread:
See these two postings for the example data set for the said 2-leg Put spread:
https://www.elitetrader.com/et/threads/calendar-secrets.369662/page-2#post-5676557
https://www.elitetrader.com/et/threads/calendar-secrets.369662/page-3#post-5676615
A quick proof ot the claimed discovery:
Assume Spot and IV stay the same at expiration like they were at entry.
Then, a LongPut will be worth $0 at expiration.
But now assume using a LongPut with a longer DTE instead. Ie. normal DTE is 60, and we use now a longer DTE 90 for the LongPut.
But still we will close the whole position on the shorter DTE 60.
Now comes the interesting part:
At the expiration of the ShortPut (ie. after 60 days) the LongPut (that did cost us 22.8528) will be worth 11.6485 (b/c it has another 30 days till its own expiration):
Pr_at_SP_expiry=11.6485 Net_Pr=11.6485 Used_Pr=11.2043 Pr_for_SP_DTE=18.0278 Saved_LP_Pr=6.8235(37.85%))
Ie. normally using a DTE 60 LongPut would have costed 18.0278, but using DTE 90 instead costs us
in the same time only 11.2043. This is 37.85% cheaper than normal (18.0278). Ie. we save that much Premium for the LongPut side of the spread. And this saving of courses reduces our cost basis, which then of course means increasing the PnL%.
Q.E.D.
PS: the calculation above is even a pessimistic one, meanng in reality the broker system recognizes (understands) that it's a spread and reduces the margin accordingly, which then of course means even higher PnL% than in the above calc.