Why Do Banks Memoryhole Your Records So Fast?
Patrick McKenzie explains ...
Say you don’t want to go through the work of submitting your income taxes on April 15, but instead ask for an extension until October 15. With my bank, I long had to remember to make sure to download my checking account records by June 30 because it only kept online 18 months of my data.
I see that they’ve now graciously upgraded to a full 25 months for CSV downloads, but still only 18 months for Quicken format. Longer than that, you have to ask to get it in printout, or something.
How come? Why can’t I look up online all my banking records for as long as I’ve been a customer?
It’s not like data storage is so expensive these days …
Patrick McKenzie, one of Stripe’s very bright Irishmen named Patrick along with co-founder Patrick Collison (see McKenzie’s take on the SPLC’s legal problems here), explains:
… here’s the math: one year to get you past next tax filing deadline, six years for retention requirement after filing for most taxpayers in absence of fraud. After that bank wants to shred it because ancient records are timebombs.
Dave Jones
@eevblog
May 25
WTF is wrong with banks, why can they only search back through 2 years of transactions?
If you want older you have to manually look through old statements, and those only go back 7 years.
And if you didn’t save older ones you are sh!t out of luck, data gone forever.
This can’t be a data space thing.
Patrick McKenzie
Why? Well nothing in a twenty year old account statement is ever good news for the custodian. Heads it does not blow up in your face.
There are an awful lot of ways to spell tails.
A really simple one is a complicated and bitter divorce or inheritance dispute. Bank wants to say “Really sorry about the situation! Best of luck.” But if it has 20 year old authoritative records it can get pulled into the problem repeatedly, for years, by opposed legal teams.
Bank wants to be able to say “Oh no we have _no earthly idea_ what happened in the joint checking account in 1977. If you have an account statement on paper, good for you! Presumably you do not need to talk to someone to explain how a checking account works, correct.”
In this circumstance, the bank is (in all probability) absolutely blameless. That is not the only thing bank legal teams are worried about in designing document retention policies, but by itself it justifies a not-infinite retention policy for basic account records.
Consider, for example, societies which went through at least one period of evil events in the past (or present). You’re welcome to pick your examples. Large society-wide evil will generally be accomplished with the use of power, and money is effectively the unit of power.
Banks keep money records and are broadly unsympathetic, and so when society-wide recriminations are getting thrown around, the person called to account tends to be less a political leader (dead/done) and less Mrs. Smith and more “whoever runs the bank with the account records.”
There exist specific examples in American and European history of banks getting made into post-hoc scapegoats for behaviors which were much broader than banks. People in charge of bank record retention policies read newspapers and history books.
And so, in expectation, the writer of a records retention policy assumes “There is a materially non-zero chance that my society will come to profoundly regret at least one thing it will unambiguously do over the course of my career or this institution’s life. In that event, we … are likely to be scapegoated. We can minimize the scapegoating by a properly scoped records retention policy, which will essentially absolve the firm on an ongoing basis for society-wide sins of the prior generation. If we fail to do this, ~unlimited potential liability.”
“Patrick are you defending banks’ behaviors in particular identifiable periods in American and European history.”
Explicitly am not. Banks are centers of societal power. You should expect them to be intimately involved in all large acts of a society, for good and ill.
But from a risk mitigation perspective nobody ever says “So we checked the history books and it turns out you guys were at ground zero for all the good news, so, we decided to pay you extra for that.”
And yeah, the particular execution of the opposite of that sentence was one-shot move in an iterated game between banks/society. The people who did it were told that, but believed the particular historical event at issue justified the one-shot.
It will be difficult to repeat it.
Anyhow: forgetting (and relatedly, forgiveness) are important technologies, and they exist in some tension with technological substrates which make perfect recall the default. Not the first nor the last time we’ve encountered that issue. Banks have kept books for a *long* time.
In illustration of McKenzie’s argument, I recall an example of a bank getting lambasted on the front page of the Wall Street Journal for allowing a client to spend the client’s own money 36 years previously. I was shocked at the time … but today it sounds more inevitable …
Paywall here.
