Modified Glass-Steagall

If you’ve read me for more than two months, you probably know that I am an actuary, though not a practicing actuary at present.  I grew up in the life insurance industry.  It’s an unusual place for an investor to start, but there are some advantages:

  • You learn some of the most complex accounting rules in business.
  • You learn the value of having a strong balance sheet, because when it slips, it is hard to get back.
  • You learn the value of simplicity, because many companies that wander from that die.
  • You get to know a lot of people with different bits of specialized knowledge, which you the actuary have to tie together.  And, respecting the older people in dead-end jobs which they do well goes a long way toward getting significant cooperation.
  • If you are a corporate actuary, as I sometimes was, potentially you become a good risk manager.
  • If are an investment actuary, you learn that risk control is far more difficult than it seems, and so you learn not to take obscure risks, and test a variety of modeling assumptions, because models can go wrong.
  • You build in margins for error if you are a pricing actuary, as I often was, and review actual results when setting assumptions.
  • You get to see regulation up close and personal, because you have to interact with 51 different regulators if you do valuation, cash flow testing, pricing, etc., with your home state regulators leading the way.

There’s more, but my topic this evening is financial regulation generally.  I’ve been thinking about it, and I have had a moderate shift in my views: I think it would be wise to reinstitute a modified version of Glass-Steagall, but modeled after the way that insurance regulation is done today.  For solvency regulation, insurers are much better regulated than banks.  The banking industry should imitate the insurance industry in a number of ways.

Here’s the main idea: Allow financial holding companies to own all manner of financial subsidiaries, but disallow:

  • Stacking of subsidiaries.  No A owns B, B owns C.  This allows capital to be stretched thin.
  • Cross-ownership and cross-lending: subsidiaries may not interlace their capital.
  • There may be no reinsurance or derivative agreements across subsidiaries.

This would bar complex ownership charts.  There would be a big box at the top, with lines to little boxes below, but only one level of depth, and no lines between subsidiaries.


  • Each subsidiary must be subject to its own regulator.  There must not be an overall regulator for “financial supermarkets.”  Keep it simple and focused.  Remember, the Fed has never been a good regulator.
  • Since financial holding companies die if they don’t get dividends, make the payment of dividends from any subsidiary to the parent company subject to the discretion of the regulators.  Regulators should not care about the holding companies, but only about the solvency of subsidiaries.
  • If a company is presently in two businesses with different regulators, the company must divide the business into two subsidiaries which each regulator can separately regulate.
  • Subsidiaries do not get to choose their regulators.  If there are potentially duplicate regulators, merge them and create one regulator if that makes sense.  Otherwise, make rules so that there is no ambiguity on who regulates what.

The view of the government toward financial holding companies should be this: we don’t care if you fail.  We do care if your subsidiaries fail, so if the solvency of any of them is getting marginal, dividends to the holding company will be cut off.

Now, I would prefer the rest of the financial industry mimic the insurance industry, in that State regulation is better than Federal regulation.  If you want to end too big to fail, split up banks into state subsidiaries.  Each state regulator would separately determine solvency issues, and would limit dividends back to the holding company.

Remember, we don’t care if holding companies go broke.  If a holding company goes broke, and all of the subsidiaries are solvent, the subsidiaries will easily be sold to other holding companies.  The creditors of the bankrupt financial holding company will divide the spoils after a year or so.  Cost to the taxpayers: zero.

And maybe, mimic the guarantee funds of the insurance industry, and let the financial subsidiaries self-fund the losses of their fallen competitors.  Cost to the taxpayers: zero.

Under this sort of arrangement, you can have “financial supermarkets,” but they would be very different, because the solvency of each part would be separately regulated.  You don’t want macro-regulators, they are far easier to fool; specialization in financial regulation is a plus; don’t give any credit to those who use a diversification argument.  We are focusing on risks, not risk.  Failure does not happen from risk in abstract, but from particular risks that were underrated.

Finally you need risk managers inside all regulated financial institutions that are either FSAs [Fellows in the Society of Actuaries] or CFAs [Chartered Financial Analysts].  I am both, though my FSA status is inactive, because I don’t pay the dues.  Why is this valuable?

You need organizations with ethics codes to teach and monitor the behavior of those within.  There are failures amid FSAs and CFAs, but society and legal punishment tends to decrease the occurrence.

If we did this, financial companies would be much more stable, and we would reduce the need for the FDIC.  There would be personal ethics standards among risk managers inside financial companies, and less reason for regulators to compromise from political pressure.

This is my modified version of Glass-Steagall, which gives financial industries most of what they want, but offers solvency protections far beyond what we have today.  Is this a good compromise, or what?