Robert Waldmann

A proposed reform (already shelved) is to require banks to offer “plain vanilla” products. I am very confused about this proposal, so this is a semi bleg. I can’t see any possible benefit from the regulation (probably because I haven’t read the fine print of the draft bill).

My thoughts after the jump.

Bottom line — a vanilla option must include the rule that at least x% of a bank’s business must be vanilla or they pay a fine to work.

Also, I propose calling non vanilla products “fudge swirl” products unless anyone has ever scene “fudge twist” or “fudge spin” ice cream.

update: Jump corrected so most of my post is after it.
Also see Waldman vs Waldmann after the jump.

I will criticize a proposal which might exist only in my imagination. A very silly vanilla rule that just says banks must offer a plain vanilla product — say a 30 year fixed rate mortgage. [Here is an] explanation by [Steve Waldman at] Interfluidity via rortybomb:

Vanilla products would turn basic financial services into a commodity business, and force providers to compete on priceā€¦. Since vanilla financial products would be commodities, banks would have to universally collude to offer them at inflated prices in order to bilk consumers. Competing vanilla project offerings would (at least they should) vary only on a single dimension (e.g. an interest rate). Points, fees, penalties, etc. would be homogeneous or uniformly pegged to the core price.

This argument does not apply to a simple requirement that banks offer vanilla products. Forcing them to offer the product does not force them to compete to sell the product.

Consider the case in which all banks offer fixed interest 30 year mortgages at 100% interest per year. Technically they have fulfilled the silly vanilla requirement. There is no improvement in anything as offering a fixed rate mortgage at 100% is just like not offering a fixed rate mortgage.

So, in the absense of collusion, is this alleged equilibrium vulnerable to deviation by a firm which offers a reasonably priced fixed rate mortgage ? It might or might not be. If it isn’t then the silly vanilla rule will not be effective. If it is, then the silly vanilla rule is not needed and will make no difference.

Since offering fixed rate mortgages at 100% is just like not offering them, the equilibrium profits and profits to deviators are just the same as in the case in which there are no vanilla products and one bank can deviate by introducing one.

If compliance with the regulation implies no real change at all, then a bad equilibrium with technical compliance will be identical to a bad equilibrium with no regulation, one is a Nash equilibrium if and only if the other is and payoffs to all agents are just the same.

So “forced to compete” only makes sense if the vanilla rule is not the silly vanilla rule. It makes sense if there is a requirement that say at least 10% of a bank’s mortgage lending must be fixed rate 30 year. Then banks will compete to issue fixed rate mortgages even if they are not as profitable as option ARMs etc , because by loaning a dollar at a fixed rate for 30 years they win the valuable right to loan 9 dollars as option arms.

I have no idea if the proposed now shelved rule was the silly vanilla rule (hence the bleg)

Steve Waldman to me
show details 5:23 AM (15 hours ago)


I tried to add this as a comment to your Angry Bear post, but alas, I am too logorrheic. The post was rejected, and I am too lazy to edit.

So I offer it to you, fwiw…


Robert — I hesitate to disagree with you, because you have that extra ‘n’ that gives you superpowers of which I can only dream.

But you are wrong, right here:

“Since offering fixed rate mortgages at 100% is just like not offering them, the equilibrium profits and profits to deviators are just the same as in the case in which there are no vanilla products and one bank can deviate by introducing one.”

This would be true if consumers had perfect information and could distinguish safe from unsafe products, vanilla from fudge swirl. But in a world where all financial products that banks offer are opaque to most consumers, and where therefore consumers attach an uncertainty cost based primarily on offering institution, no bank has an incentive to deviate with a less profitable vanilla product. Suppose a 30-year “fudge swirl” mortgage creates higher revenues in expectation (and higher costs to consumers) than a 30-year vanilla. But consumers cannot distinguish fudge swirl from vanilla. Then offering only fudge swirl is the dominant strategy for banks. Offering vanilla involves fixed costs of product development, and consumers that choose vanilla are just an opportunity cost viz fudge swirl.

Now suppose a trusted external party, call it “the government” certifies some products as vanilla. (There is no other party that could credible on this given the incentives to game certifications, and even the government might be too compromised.) Now consumers can distinguish between vanilla and fudge swirl, and deviating involves benefits as well as costs. Offering a non-100% 30 yr vanilla will capture some extra consumers, who know it is that good clean flavor they like, leading to extra revenues, while it will also cannibalize some existing, more profitable nonvanilla prospects. Under this circumstance, there should be deviation, as the sole provider of vanilla (under reasonable assumptions) would gain much more by taking other firms’ prospects than they’d lose downselling their own client base.

This argument suggests that there’s no reason to require any bank to offer vanilla, just a requirement to have the government certify some products as vanilla. And I think that’s mostly right, although I support a requirement for the sake of timing, because it can take a while for existing well-policed cartels to collapse despite a favorable Nash equilibrium. That is, if CFPA simply defined a schedule of vanilla financial products and offered to certify compliant offerings, the larger banks would “as a matter of principle” refuse to participate, and many smaller banks as well by virtue of industry/ABA pressure. The industry would also lobby for very elaborate certification requirements for vanilla products, and that banks should pay their own certification costs, to create a barriers that would help to discourage smaller deviators. Still, if the certification barriers aren’t too outlandish, some small and midsize banks would deviate. Since financial product markets are segmented (there are many customers who for arguably misguidedly perceive larger providers as safer or more reliable), big banks wouldn’t be under very much pressure to follow suit and compete. Eventually, again if certifiation costs aren’t outlandish, the cartel would break, and fundamentally, mere credible certification of vanillahood would be sufficient over the long run. But over the long run we’re dead. I support vanilla as a requirement because I want to see informationally-protected financial services rents collapse quickly, not “at equilibrium”.

Note that the market that would be least affected by vanilla products is mortgages, because a de-facto vanilla standard already exists — GSE requirements for prime mortgages. Here we can see that…

1) vanilla is not a panacea — there has long existed a widely known vanilla product, yet banks often persuaded prime-eligible consumers to accept products with much higher expected costs by offering attractive features like teaser rates and negative amortizations. People can still choose belly-ache when vanilla is on offer, and they will.

2) vanilla is not useless. Many homebuyers, understanding that mortgages are a complex game they are ill-equipped to play, forgo lower down- and monthly payments to stick with “traditional prime” mortgages. Vanilla has worked exactly as it is supposed to for this population: homebuyers can shop around for a prime mortgage based on fees, perceived service, convenience, etc, and the market is competitive. Not all providers are the same, of course, but people are rarely screwed by surprising characteristics of the product. Lots of primes are defaulting of course, but they are doing so for reasons they would have understood well ex ante: they can’t make the monthly payment, even though it is the same payment they signed up to and paid their first month. They are not being screwed by prepayment penalties, resets, recasts, interest-rate spikes, etc. That primes are suffering in the quantity that they are is arguably a function of the popularity of nonprime products, which permitted higher effective leverage and therefore helped inflate a housing market primes had to compete in. (If all homebuyers had to put 20% down per supervanilla prime, there would have been no housing bubble.)

Anyway, my understand is that the vanilla rule would have been what you are calling the “silly vanilla rule”. But you are wrong to do so. Such a rule would not be silly at all.