Traditional accounting starts with a fairly comforting assumption: the financial records were designed to help somebody understand the finances.
Payments frequently skipped that meeting.
A single merchant transaction can generate interchange, network assessments, processor fees, gateway fees, sponsor-bank economics, ISO residuals, reserves, chargebacks, ACH movements, adjustments, and several other entries whose descriptions appear to have been written by someone with a personal grudge against vowels.
Multiply that by millions of transactions, multiple processing platforms, settlement accounts, FBO structures, and years of activity, and apparently simple questions become surprisingly difficult.
How much money should be in this account? Whose money is it? Why is the account short? What did the processor actually charge? Did the residual calculation change? Are interchange costs being passed through correctly? Is that a real card-brand fee, or did somebody invent a fee and give it a sufficiently official-sounding name?
We have spent years working inside these systems. We know the files. We know the economics. We know the weird reports. And, perhaps most importantly, we know which numbers should reconcile even when the systems producing them seem determined to prevent that from happening.
This is not ordinary accounting. It is closer to financial archaeology with significantly more CSV files.