- Critical Events (purchase of Bear Stearns and WaMu and acceptance of TARP funds).
- "Blame" for the crisis (possible root causes for it).
- Future and regulation.
Summarizing his letter here would be pretty hard, since it contains a lot of topics and is highly summarized already. So I will focus on two interesting aspects and possibly the most controversial ones: assigning blame and more regulation.
He starts with a caveat/disclaimer:
The causes of the financial crisis will be written about,And then proceeds to enumerate them:
analyzed and subject to historical revisions for decades.
Any view that I express at this moment will likely
be proved incomplete or possibly incorrect over time.
• The burst of a major housing bubbleThe Burst of housing bubble
• Excessive leverage pervaded the system
• The dramatic growth of structural risks and the unanticipated damage they caused
• Regulatory lapses and mistakes
• The pro-cyclical nature of virtually all policies, actions and events
• The impact of huge trade and financing imbalances on interest rates, consumption and speculation
Poor underwriting standards (including little orExcessive leverage
no verification of income and loan-to-value ratios as
high as 100%) and poorly designed new products (like
option ARMs) contributed directly to the bubble and
its disastrous aftermath.
He cites hedge funds, banks (and their SIVs), private-equity firms, investment banks, CDO managers, some bond funds and even pension plan funds were highly levered, due to conditions made possible from rising prices, thus fueling even higher prices. He concludes:
Basically, the whole world was at the party, high onThe dramatic growth of structural risks and the unanticipated damage they caused
leverage–and enjoying it while it lasted.
He starts with short-term, illiquid assets:
Some banks, hedge funds, SIVs and CDOs were using short-term financing to support illiquid, long-term assets. When the markets froze, these entities were unable to get short-term financing. As a result, they were forced to sell these illiquid assets.And goes on to money market funds:
After Lehman collapsed, one money [market] fund in particular, which held a lot of Lehman paper, was unable to meet the withdrawal demands. As word of that situation spread, investors in many funds responded by demanding their moneyAnd then on to repo financing (daily borrowing between banks):
[I]n those markets where financial companies financed their liquid assets, financial terms had become too lax. For example, to buy non-agency mortgage securities, financial institutions only had to put up 2%-5% versus a more traditional 15%-25%. The repo markets also had begun to finance fairly esoteric securities, and when things got scary, they simply stopped doing so. [...] Once again, financial institutions had to liquidate securities to pay back short-term borrowing.Regulatory lapses and mistakes
Again, he starts with a disclaimer:
With great hesitation, I would like to point out that mistakes also were made by the regulatory system. That said, I do not blame the regulators for what happened.Then, on to lack of regulation or poor regulation throughout the system:
Much of the mortgage business was largely unregulated. While the banks in this business were regulated, most mortgage brokers essentially were not.He continues:
Basel II [...] was highly flawed. It was applied differently in different jurisdictions, allowed too much leverage, had an over-reliance on published credit ratings and failed to account for how a company was being funded (i.e., it allowed too much short-term wholesale funding).And on Fannie and Freddie regulators:
Perhaps the largest regulatory failure of all time was the inadequate regulation of Fannie Mae and Freddie Mac. [...] Both had dramatically increased their leverage over the last 20 years. And, amazingly, a situation was allowed to exist where the very fundamental premise of their credit was implicit, not explicit.The pro-cyclical nature of virtually all policies, actions and events
This is possibly the most interesting part of Dimon's recount of why the crisis unfolded. The high-level view here is that when things are bad, regulation and accounting rules force increased loan-loss reserves (thus consuming capital), mark-to-market accounting distorts the fair value of illiquid assets, rating downgrades can force some insurers to post more collateral (thus also burning through capital), financing arrangements (loan covenants, etc) force deleveraging at the worst possible times, while allowing higher leverage in good times, and with increased volatility Basel II accounting rules demand more capital be held.
In a crisis, pro-cyclical policies make things worse. I cannot think of one single policy that acted as a counterbalance to all of the pro-cyclical forces. Although regulation can go only so far in minimizing the impact of pro-cyclical forces in times of crisis, we still must be aware of the impact they have.The impact of huge trade and financing imbalances on interest rates, consumption and speculation
Over an eight-year period, the United States ran a trade deficit of $3 trillion. This means that Americans bought $3 trillion more than they sold overseas. Dollars were used to pay for the goods. Foreign countries took these dollars and purchased, for the most part, U.S. Treasuries and mortgage-backed securities. It also is likely that this process kept U.S. interest rates very low, even beyond Federal Reserve policy, for an extended period of time. It is likely that this excess demand also kept risk premiums (i.e., credit spreads) at an all-time low for an extended period of time. Low interest rates and risk premiums probably fueled excessive leverage and speculation.And this ends his summary of the root causes and main fuel to the great financial fire in history. The next part is his proposal for how to avoid another one going forward or at least how to minimize its scale and impact. It is probably a very controversial topic as it hinges on more and better regulation, which is typically a double-edged sword.
The extent of the damage and the magnitude of the systemic problems make it clear that our rules and regulations must be completely overhauled.And cautions:
Political agendas or simplistic views will not serve us well.Then:
Often we hear the debate around the need for more or less regulation. What we need is better and more forward-looking regulation.The plan:
The first goal should be to regulate financial institutions so they don’t fail. If they do fail, a proper resolution process would ensure that action is swift, appropriate and consistent. The lack of consistency alone caused great confusion in the marketplace. For example, when some of the recent failures took place, there was inconsistent treatment among capital-holders (preferred stock and debt holders were treated very differently in different circumstances). It would have been better if the regulators had a resolution process that defined, a priori, what forms of aid companies would get and what the impact would be on capital-holders. The FDIC resolution process for banks provides a very good example of how a well-functioning process works.And on fragmentation of the many regulators and their limited scope:
If [...] similar products were overseen by a single regulator, that regulator would have much deeper knowledge of the products and full information that extends across institutions. [...] Everyone agrees that the existing system is fragmented and overly complex. We have too many regulators and too many regulatory gaps. No one agency has access to all the relevant information. Responsibility often is highly diffused.And hedge funds:
[Hedge funds] could be required to show their regulators (not their competitors) any concentrated “trades” that could cause excessive systemic risk.Mortgages:
Mortgages are the largest financial product in the United States, and while we do not want to squelch innovation, the entire mortgage business clearly needs to be regulated.Securitization:
No one was responsible for the actual quality of the underwriting. Even mortgage servicing contracts were not standardized such that if something went wrong, the customer would get consistent resolution. We cannot rely on market discipline (i.e., eliminating bad practices) alone to fix this problem.On fixing mark-to-market (or "fair value accounting") he says:
We generally like fair value accounting.But provides a caveat when times are bad, which requires companies to use synthetic benchmarks and judgment on "comparable" securities:
In many instances, cost is the best proxy for fair value. We would rather describe our investments to our shareholders, tell them when we think these investments might be worth more and, certainly, write them down on our financial statements when they have become impaired.Which is interesting if it's not abused. But when there's leeway for good faith estimates, it's bound to be abused. I've commented on closed-end mutual funds estimating the value of treasuries based on other methods besides market values.
He essentially agrees with Charlie Munger on what I called "voodoo accounting":
Essentially, we now have to mark to market credit spreads on certain JPMorgan Chase bonds that we issue. For example, when bond spreads widen on JPMorgan Chase debt, we actually can book a gain. Of course, when these spreads narrow, we book a loss. The theory is interesting, but, in practice, it is absurd. Taken to the extreme, if a company is on its way to bankruptcy, it will be booking huge profits on its own outstanding debt, right up until it actually declares bankruptcy–at which point it doesn’t matter.And then goes on:
It is becoming increasingly more difficult to compare mark-to-market values of certain instruments across different companies. [...] [D]ifferent companies may account for similar mark-to-market assets differently. This needs to be addressed by ensuring that companies adhere to consistent valuation principles while applying the rules.Not all illiquid assets need special treatment (such as real estate, plant and equipment), but some others do. He cites examples:
Fair value accounting does not and should not apply to all assets. [...] [F]or example, a farm would be worth more when corn prices go up, and a semiconductor plant would be worth less when semiconductor prices go down. [...] [M]arking the aforementioned assets to market every day would be a waste of time. Under this scenario, it would be quite hard for companies to invest in anything illiquid or to make long-term investments.And now perhaps the most interesting part is what this new regulation and accounting method should do, to counter what he previously called "cyclical policies". He cites three good ideas:
- Loan loss reserving can easily be made counter-cyclical
- Repo and short-term financing can easily be made counter-cyclical
- Banks should have the ability to implement counter-cyclical capital raising with rapid rights offerings
As things get worse and charge-offs rise dramatically, one must dramatically increase loan loss reserves, thus depleting capital rapidly. This problem would be solved if banks were allowed to estimate credit losses over the life of their loan portfolios. Reserves should be maintained to absorb those losses. This would enable banks to increase reserves when losses are low and utilize reserves when losses are high. Transparency would be fully preserved because investors and regulators would still see actual charge-offs and nonperformers. This would require a rational explanation about the appropriateness of the lifetime loss estimates. It also would have the positive effect of constantly reminding CEOs, management teams and investors that bad times, in fact, do happen–and that they should be prepared for such events.Repo and short-term financing
If an institution provides financing to clients in excess of what the Fed would lend to the bank for the same securities, it would have to be disclosed to risk committees and the company’s Board of Directors. The Fed then would have two major tools to reduce leverage and in a way that is counter-cyclical–it could charge higher capital costs to a bank when the bank is lending more than the Fed would lend or the Fed could reduce the amount it would lend to the banks.Rapid rights offerings
Banks and possibly other companies would be aided by having the ability to effect rights offerings at a moment’s notice. Regulations should facilitate such offerings–with the proper disclosure–in a matter of days rather than weeks.
All told, it was a very interesting read. The summary of the historical facts paint a very complex picture. Saying it was regulators' fault or the banks' fault or even subprime borrowers' fault is too naive. It was a combination of all these things over a long time that led to excesses that had to be undone. We're now paying the price for letting it happen.
Perhaps the most controversial points were his proposed solutions. More regulation can be a bad thing. I'm a believer that "free" markets are one of the most powerful forces in nature. On the other hand, I don't believe that free means "free of regulation". I believe free means without government intervention in prices and the free exchange between market participants.
For this free market to happen, however, there needs to be a healthy dose of rules -- regulation, if you will. Markets without rules and penalties for those who violate it are bound to be exploited by evil participants, which will cause others participants to take measures that will defeat the benefit of the free market and impose too high a cost to smaller participants. Ultimately, unregulated markets will fail.
Striking a reasonable balance is necessary. Dimon's suggestions did not seem unreasonable. But at the same time, they didn't seem to be that powerful. I don't think they would resolve the problem going forward, though they would certainly help. I'm still undecided about the idea of letting banks tell us what the fair value of their assets are. I'd be more convinced when I knew what the details and guidelines were and what kind of punishment was there for those who abuse it -- and how regulators plan to audit and detect abuses.