Nostradoofus

The previous post on credit card issuer default rates deliberately avoids the question of ethics in lending.  I’ll mostly evade that again here, except for one point:  it would seem patently unethical to earn de facto interest in ways that are not readily understandable to the borrower.

By that standard, it would not necessarily be unethical to loan someone $100 revolving at 40% interest.  I give you $100 today, and you give me $140 next year, or pay $40 next year and roll it over to the following year, etc.  Ridiculously expensive, but explained up front, and readily understandable to the borrower.

But what if I instead lend you $100 at 5% interest, and buried in a long, complicated list of loan terms, in tiny print, is a fee of $30 for late payment, plus a 30% “default rate” that applies retroactively to the beginning of the loan in the event of late payment?  What if I then send no reminders?  Suddenly the effective rate is 60% for anyone who pays the first year’s interest just one day late.

A deal’s a deal.  Yes, it is.  But there is also a “reasonable man” provision in contract law, one I think has not received adequate attention in recent years.  It may be reasonable to expect someone to understand the consequences of 40% interest, but unreasonable to expect someone to understand and remember the dozens of complex features of the dozens of financial products in his life.  

We may have reached, or long passed, the point at which the complexity of personal finance exceeds the capacity of the reasonable man, who then by law cannot be considered negligent if he fails to understand something.  

Watch for something like this in coming years.  Nostradoofus has spoken.

—

I’ve mentioned before the general principle that optimization creates inflexibility — in business, investing, computer programming, law and much else.

The credit card industry’s current travails offer an example.  Many have written on this subject recently, many good, but usually with the leftist or moralist subtext that 30% interest rates are somehow inherently wrong, so card issuers are getting what they deserve.  I’ll leave the Biblical debates to other writer.  This post just talks about the relationship between steady-state profit maximization and brittleness in economic shocks.

Credit card issuers mainly make money from so-called “revolvers” — people who endlessly pay interest on high balances, too high to pay down, but never quite high enough to trigger  default.  The issuer’s business revolves around revolvers, endlessly optimizing to maximize cash extraction from that particular group.  (Conceptually similar to the way casinos make much of their money by catering to and optimizing for “whales,” a gaming industry term for high-stakes compulsive gamblers.)

The 2005 bankruptcy reform, in force since 1/1/06, was intended as yet another credit card industry optimization.  By making Chapter 13 harder to qualify for, debtors could not easily write off consumer even in bankruptcy.  Instead, they were placed in repayment programs by the bankruptcy court.  According to at least one reliable source, issuers responded to this by lending more freely, assuming they would always be able to collect, even in bankruptcy.

This and other optimizations were in a sense too successful:  the more effectively the issuer optimizes, the closer it pushes the “revolver” to his absolute theoretical fiscal limit, the more exposed both borrower and issuer become to an economic shock.

Say you are a lender.  Would you want most of your customers to be paying you 100% of their EBIT in interest payments?  A residential lender would answer no.  Corporate bond issuers would say “no way.”  Too dangerous, no margin of safety.  A one-dollar decrease in the borrower’s income would trigger default.  Yet card issuers were actively seeking that 100%.  They tried to control losses with higher rates and better collections, but mainly just crossed their fingers on general economic stability.  That turned out to be an ill-placed hope.

—

It’s interesting, initially tragic and eventually uplifting, to watch non-technical friends and relatives use Mac after years on Windows.  Like rats who have been shocked at random intervals in the lab, they exhibit persistent learned helplessness at the keyboard, fearing to tread outside the few simple use patterns they are familiar with.  When presented with a new situation, they simply freeze.

But I have developed a solution, one that could only work on Mac.

“How do I get my computer to do X?” I am asked at regular intervals.  If the inquirer is on Windows, I give them a lengthy step-by-step answer, involving 12 levels of nested menus and dialogs, right clicks, genuflecting, etc.

But if they are on Mac, even when I know the answer, I first ask, “What do you wish were the answer to that question? What would be the most logical, dead-simple, effortless answer?  Try that.”  Nine times out of ten, it works.  After this happens a few times, the user is unstuck, and can enjoy computers for the first time.

—

My favorite piñata for demonstrating the principal-agent problem is Microsoft marketing.  But Time Warner also offers good examples in which long-run earnings are sacrificed to serve short-run interests of individual ladder-climbers.

Last year, my kids received the entire Looney Toons collection from a generous uncle.  This vast DVD set includes nearly every Warner Brothers animated short from the 1930s to 1960s.  The kids absolutely loved it, because — and I didn’t yet realize the significance of this — they had never seen Bugs Bunny or Daffy Duck before.

My daughter, then 7, invited a friend over to see.  ”Let’s watch Daffy Duck,” she said.  ”Who?” replied her friend.  It was hard to get her interested, because, I finally realized, no one under 8 in this area has ever seen Bugs or Daffy before.  They are not broadcast here.

You can see how such a situation might arise.  The DVD collection is one-of-a-kind, exhaustive, and very expensive.  The people who buy such things are over 30 and not price sensitive.  So in the short run, if you make the cartoons unavailable on TV, enthusiasts are more likely to buy the DVD.  Sales go up.  For now.

But the TV shows drive DVD demand 20 years on.  In a couple of decades, DVD sales (more likely their digitally delivered equivalent) will collapse, as the first crop of parents arises who have never heard of Bugs and Daffy.

By then, the marketing strategist who came up with this bright idea will be long gone, perhaps even promoted for his ability to increase short-term contribution.  That guy is not necessarily evil — he may simply have responded to bad incentives.  Designing those incentives is the hardest problem in management.

—

The Fewer, Nicer Things meme applies in particular to government.  Minimizing regulatory compliance expense is mainly a matter of simplifying complex, capricious, self-contradicting rules.

Do the endless, ever-shifting sands of TSA flight regulations actually make you safer?  Unlikely, because planes are too complex to survive a determined terrorist.  Instead, millions of people are inconvenienced, and billions of dollars in time and money wasted, for essentially political reasons:  so politicians can look like they’re helping.  Much of the inconvenience is in spending valuable time and mental effort to figure out what the rules are.  Even the enforcement officers can’t seem to figure it all out.  Not their fault — it’s just too complicated.

A well-designed, extremely simple regulatory regime would solve this.  For example, “Your carry-on must contain no metal, and must fit in a cigar box.”  This is an extreme pain, and still does nothing to make you safer, but is extremely easy to understand, comply with, and enforce.  Baby steps back toward sanity.

Thomas Jefferson argued that the entire body of federal laws should be thrown away and rewritten every 30 years, to keep them simple, current and reflective of the public will.  That’s extreme (hey, Jefferson was kind of an extremist), but a related idea in the Jeffersonian mold would be this simple Constitutional amendment:

“Every year for the next half-century, the US legal code shall contain 1% fewer words than the year before, and no congressman shall vote on a law he has not personally read in its entirety.”

This is a very simple way to limit complexity.  You could refine this rule, of course, but that would make it complicated.  The goal is simplicity.  Is this realistic?  Very.  Much of the legal code is obsolete, redundant, or consists of narrow exceptions to and refinements of existing law.  (As programmers would say, they have been applying patches and bug fixes, when they should be refactoring the code.)  Moreover, in the past 30 years, much law was written with almost no regard for complexity.  For example, the securities regulations of 1933-40 were relatively simple, but have been vastly expanded since 1970 by exceptions and refinements — apparently without effect on our financial safety.

Simple laws — as long as they are designed for measurability — require less time and effort to write, read, interpret and enforce, and thus make the entire system more efficient.

—

Yet another example of the Orphans of the Sky scenario at Microsoft is Google’s use of Microsoft’s ActiveSync to synchronize your contacts and calendars among your mobile and cloud-based apps.

In case the buzzwords are unfamiliar:  Google provides free contact management and calendar functions through Gmail, its free email service.  ActiveSync is the data synchronization component of Exchange, Microsoft’s extremely expensive server application for contact management, calendar and email.  So one is free, the other extremely expensive.  Got that?  Let’s move on.

It is really hard to make synchronization like this work right.  This was Palm’s secret sauce for a while, and they still hold a bunch of patents on it.  Mighty Apple Inc. blew it here recently, accidentally erasing all contact data for thousands of customers during a routine software update.  Microsoft no doubt spent a lot of money and effort getting ActiveSync just right.  But they apparently forgot their most valuable skill of the 1980s and 1990s:  strategy.

As Palm learned to its disappointment, there is little strategic advantage in replicating data across devices.  The real advantage is in being the place where that data is generated and stored.  

If you use Gmail and its contact and calendar apps, then Google holds the strategic advantage (albeit a weak one, because everything can be exported to open formats).  They paid almost nothing for the synchronization features.  Microsoft paid a lot for synchronization features that actually facilitated their customers’ transition to Gmail.

If you synchronize via Exchange, but maintain all your contacts at Google Apps or Gmail, then Google has won another round against MSFT.

—

Many inveterate readers I know have stopped reading their usual news.  The cause seems to be crisis fatigue.  Readers are going fetal.  

This points to an interesting instability in psychology:  up to a point, media gets more attention and credibility with vivid negative stories;  beyond that point, readers and viewer simply disengage.  The transition is abrupt, so it surprises industry experts when it happens.

Same thing seems to have happened in politics in the past year.  Negative, divisive campaigns, and fear of terrorism, nearly always work.  Except this time.