Makes sense, although I still like the idea that the final message was:
and now we need to ... Subscribe to Dan Davies - "Back of Mind"
Which probably would take a little more warranting than is present in the post as drafted, but is doubtless good advice for the SEC, the FCA, and anyone else involved in regulation in this context.
From a previous commenter: «From the private banker's perspective, a compliance officer's job is to stay out of the way and serve as fall guy if the bank is caught. At least in America, it's very hard for a person making $250,000 to push back against a person in the same firm making millions.»
Especially if the person on 250k is a cost and their job blocks profitable deals and the person on millions is a rainmakes and their job makes profitable deals.
«a simple statement which everyone will disagree with – the compliance officers and the regulators are, despite everything, working on the same team.»
They have the same *purpose* but they are on different teams: the pay and promotions of the compliance people and of the regulators are decided by entirely different organizations according to entirely different metrics.
Perhaps I am not a sophisticated cybernetician like our blogger, but let me reason from a position of naivety, one that looks more at incentives and interests than information flow engineering:
* In some regulatory systems the compliance checking is done by teams of regulators embedded within the regulated entities, in which case the compliance checkers and the regulator are literally the same team, they do not have just the same ostensible purpose.
* In that case the interests of the compliance officers would be to focus on important cases and less on covering their backs by spamming insignificant cases, and their incentive would not be to avoid offending the traders on which indirectly their pay and promotions depend (but of course there are "chinese walls" so clearly I do not understand how the world works), but please their enforcement colleagues who write their performance assessments.
Instead it has been for some decades quite popular (not just for AML not just in finance) for government to mandate self-certification for regulated organizations, by asking them to hire their own compliance checkers which has some awesome consequences:
* The self-checking staff gets a huge interest in pleasing the top executives some way or another and a great incentive to obsess about a large number of small issues.
* Being an unfunded mandate it means that taxes need not be raised to fund the costs of compliance checking, and those costs will be shifted by the self-certifying organization onto their customers with the weakest negotiating power.
Look like a win-win to me!
«the problem of AML regulation, but we can now at least see that it’s basically an engineering problem, where we have made a decision to put an important piece of information processing capability on one side rather than another of an important attenuator»
Perhaps I am too dumb to fully understand it is “basically an engineering problem” and not a *solution* that aligns with the interests and incentives of the people who matter as I naively was guessing.
What would you identify as the system 5 philosophy / identity function in these cross-boundary organizations? It seems to me that one way to describe the difference between the AML problem and the digital-asset problem is that (1) you can argue that there is a shared, constant-over-time definition of what money laundering is and why it's bad, with disagreements basically being over how much time and effort and opportunity costs (in the form of legitimate business that gets flagged as suspicious and therefore not done), but (2) it is much harder to argue that there is a shared, constant-over-time definition of whether, when, and how transactions involving digital assets should be regulated as securities issuances or transactions, and if so what if any changes should be made to existing regulations because of their different characteristics, and so forth. And it seems to me the latter is a system 5 problem because it requires us to know who "we" are and why we care about what we're doing here.
... change other things to accomodate that decision.
Was the missing last line - I would edit the post but I am always paranoid that to do so would result in the email being sent again!
Makes sense, although I still like the idea that the final message was:
and now we need to ... Subscribe to Dan Davies - "Back of Mind"
Which probably would take a little more warranting than is present in the post as drafted, but is doubtless good advice for the SEC, the FCA, and anyone else involved in regulation in this context.
Shucks, I thought it was such a brilliant way do end, too (not that I'm criticizing the real ending or anything).
"way TO end" I mean.
From a previous commenter: «From the private banker's perspective, a compliance officer's job is to stay out of the way and serve as fall guy if the bank is caught. At least in America, it's very hard for a person making $250,000 to push back against a person in the same firm making millions.»
Especially if the person on 250k is a cost and their job blocks profitable deals and the person on millions is a rainmakes and their job makes profitable deals.
«a simple statement which everyone will disagree with – the compliance officers and the regulators are, despite everything, working on the same team.»
They have the same *purpose* but they are on different teams: the pay and promotions of the compliance people and of the regulators are decided by entirely different organizations according to entirely different metrics.
Perhaps I am not a sophisticated cybernetician like our blogger, but let me reason from a position of naivety, one that looks more at incentives and interests than information flow engineering:
* In some regulatory systems the compliance checking is done by teams of regulators embedded within the regulated entities, in which case the compliance checkers and the regulator are literally the same team, they do not have just the same ostensible purpose.
* In that case the interests of the compliance officers would be to focus on important cases and less on covering their backs by spamming insignificant cases, and their incentive would not be to avoid offending the traders on which indirectly their pay and promotions depend (but of course there are "chinese walls" so clearly I do not understand how the world works), but please their enforcement colleagues who write their performance assessments.
Instead it has been for some decades quite popular (not just for AML not just in finance) for government to mandate self-certification for regulated organizations, by asking them to hire their own compliance checkers which has some awesome consequences:
* The self-checking staff gets a huge interest in pleasing the top executives some way or another and a great incentive to obsess about a large number of small issues.
* Being an unfunded mandate it means that taxes need not be raised to fund the costs of compliance checking, and those costs will be shifted by the self-certifying organization onto their customers with the weakest negotiating power.
Look like a win-win to me!
«the problem of AML regulation, but we can now at least see that it’s basically an engineering problem, where we have made a decision to put an important piece of information processing capability on one side rather than another of an important attenuator»
Perhaps I am too dumb to fully understand it is “basically an engineering problem” and not a *solution* that aligns with the interests and incentives of the people who matter as I naively was guessing.
It's "le" système "financier" though. (Unless that's a trick.)
What would you identify as the system 5 philosophy / identity function in these cross-boundary organizations? It seems to me that one way to describe the difference between the AML problem and the digital-asset problem is that (1) you can argue that there is a shared, constant-over-time definition of what money laundering is and why it's bad, with disagreements basically being over how much time and effort and opportunity costs (in the form of legitimate business that gets flagged as suspicious and therefore not done), but (2) it is much harder to argue that there is a shared, constant-over-time definition of whether, when, and how transactions involving digital assets should be regulated as securities issuances or transactions, and if so what if any changes should be made to existing regulations because of their different characteristics, and so forth. And it seems to me the latter is a system 5 problem because it requires us to know who "we" are and why we care about what we're doing here.