"You've got me? Who's got you?" -- Lois, when assured by Superman during her free fall, in what I consider the epic of critical thinking.
The Globe And Mail had a report on a fascinating proposal by Canadian Finance Minister:
Finance Minister Jim Flaherty has indicated on several occasions that his counterproposal of forcing banks to sell debt that would convert to equity in times of stress, or contingent capital, is more than a bargaining ploy.
I haven't been able to find any details of it. But I'm guessing it's along the line of reversible convertible bonds (RCB) with a twist. If so, it'd be so much better than anything that's been thrown around in the US Congress or the G-20 meeting, or anywhere else I've seen, ideas which are all dead on arrival since they all involve direct government bailout one way or another.
This approach forces TBTFs to be self-reliant, with minimum government intervention (only setting up the requirements and enforcing them) and letting the market decide how much each bank should pay for its own insurance.
Here's my version of the approach:
Banks are required to sell RCBs, the value of which after conversion must be no less than a percentage of their total on- and off-balance sheet liabilities at any time, end-of-day basis. Banks decide the maturities and conversion dates/prices/ratios as long as there's no gap in conversion periods. The market sets the coupons. If times are good, it's just like a regular bond (but with higher coupon). If the bank is in trouble and shares plunge, it can convert the RCB into equity at pre-set prices to replenish balance sheet (lower liability and more equity) and meet capital requirements.
It's costly to banks. And that's the point. Banks have many flexibilities in how to structure it. But each option comes with a price and this is also the point. Shorter maturity/conversion dates give the banks lower coupon amounts to pay for; but they run the risk of having to roll it at inconvenient times. Lower conversion prices give the banks lower premium to pay for; but they'd have to sell more of it to meet the minimum requirement.
But there should be an early redemption option. If the bank has excess RCB and feels confident it will not need it for a long time, it can redeem them early. Frequent early redemption and selling are costly and damaging to confidence. This forces banks to have a clear, stable, long-term strategy and encourages them to control risk. Any breach of the minimum requirement would be met with stiff penalty.
Here's why it could work:
1. It's better than Fed reserve requirement.
It's a disaster insurance that banks must buy, not from any single counterparty, which of course doesn't work (as proven by AIG), but rather from the market. Fed reserve has proven completely inadequate in the post-Glass-Steagall era, because the requirement is based on depository base. This approach is based on total liability. If banks don't like it, they can always support re-enacting Glass-Steagall and just deal with Fed reserve requirement.
2. It's better than tax.
The bank tax idea is so dumb I suspect whoever came up with it did it for the specific purpose of letting it fail. Every bank is different. What do you base the tax on? If bank A makes money on old fashioned loans while bank B does it through prop trading, is it fair for them to pay the same rate for this special tax? Furthermore, the tax implies government would have to bailout although the tax income before the crisis will have been spent years in advance. It is so wrong at so many levels I'll just stop here.
3. It's better than arbitrarily limiting leverage.
Limiting banks' leverage is also a dumb idea because the "reasonable" limit, even if there exists such a thing, will have to heavily depend on the assets and trading strategy. Who is to say what that magic number should be? This approach, in contrast, puts a flexible limit on leverage that's set by the market according to the changing conditions of the banks. It does have a pre-set number, the percentage of post-conversion value relative to total liability. But it's up to the market to judge the quality of the balance sheet and demand yields accordingly.
4. It's even better than Volcker Rule or re-enacting Glass-Steagall.
Re-enacting Glass-Steagall is probably impractical given today's reality in finance. Defining what is prop trading or "proper" use of derivatives can be very tricky, thus susceptible to manipulation and abuse. With this approach, banks have the choice of what to do with their business. If they want to go back to old fashioned boring commercial banking, leverage will be very low and cost of this insurance low. If they want to go all the way out, fine, as long as they can make enough to pay for the insurance.
5. It's better than raising equity during crises.
Raising equities during a crisis is often suicidal because of two factors: dilution and perceived desperation. This approach still dilutes; there's no way around that, nor should there be. But it significantly reduces the shock. As a bank's situation deteriorates and shares slide, the prospect of RCB conversion is there for all to see. People know the amount of extra equity well in advance, even the time as the share price approaches the conversion points. The bank can structure the insurance RCBs such that there'll be a series of knock-in stops. Unless the bank is beyond hope, the share price will not go into a death spiral upon conversion because the conversion strengthens its financial soundness. It will not cause the feedback loop such CDS does because it automatically stops somewhere near the conversion points.
5. It provides unprecedented transparency to banking risk.
With periodic government audit and threat of heavy penalty, banks are forced to keep a safety margin on the amount of excess RCB and adjust it (selling more, redeeming early, or not rolling certain tranches) according to their own situation. This provides the market extremely valuable insight into the banks' balance sheets and management's plans.
Banks won't like this. But given the alternatives, I believe this is by far the least of all evils. If I were a bank, I'd embrace this with enthusiasm and focus my lobbying army on lowering the minimum percentage number. It's up to the society at large to demand a minimum on that minimum.
It doesn't solve all problems. Nothing does. There's still a risk that a bank remains insolvent even after all such RCBs convert to equities. If the minimum requirement is so high as to substantially eliminate such a risk, it'd probably kill banking altogether. But along with forced transparency in derivatives exposure, this should go a long way toward reducing systemic risk with the huge advantage of minimal, arbitrarily set limits and government intervention.