Four weeks ago, Treasury Secretary Geithner announced the administration’s interest in developing a plan—which the Treasury is willing to back with up to $1 trillion of public funds—to partner with private capital to buy banks’ “troubled assets.” The announcement has met with substantial skepticism about the possibility of working out an effective plan to restart the market for troubled assets for such a public-private partnership. However, in a paper issued last month, How to Make TARP II Work, I show that this can be done, and explain how the plan should be designed to contribute most to restarting the market for troubled assets at the least cost to taxpayers.
The Bush administration was forced to abandon its own plan for directly purchasing troubled assets once it became clear that its plan could not be designed to effectively address concerns about arbitrary valuation and potential overpayments. This experience contributed to the doubts about the Obama administration’s new plan. With the stock market reacting negatively, one noted columnist observed that “[t]he market was right to worry because… nobody has yet devised a way to make such a scheme work.” Another columnist suggested that the market was “glum” because the announcement was “short on details – and no more so than on the critical question of how the government will address the problem of dealing with the toxic assets that have effectively rendered large portions of the nation’s financial system insolvent.”
Despite the widespread doubts, an effective plan for a public-private partnership in buying troubled assets can be designed. The key is to have competition at two levels. First, at the level of buying troubled assets, the government’s program should focus on establishing many competing funds that are privately managed and partly funded with private capital – and not creating one, large “aggregator bank” funded with public and private capital and engaging in purchasing troubled assets. Second, at the level of allocating government capital among the competing private funds, potential fund managers should compete for government capital under a market mechanism resulting in maximum participation of private capital and minimum costs to taxpayers.
The premise of the Treasury’s plan is that the banks’ current problems are at least partly due to the freezing up of the market for many kinds of troubled assets. The concern is that money managers otherwise willing to purchase such assets at any price levels below fundamental value do not have sufficient liquidity to keep prices from falling below such levels. The Treasury therefore seeks to restart the market for troubled assets by introducing sufficient additional capital on the buying side of this market.
To bring additional capital to the market’s buying side, the Bush administration initially planned to spend $700 billion on direct purchases of troubled assets, only to abandon the idea amid concerns about valuation of such assets. The Obama administration seeks to get around the valuation problem by involving private parties—viewed as better positioned and incentivized to make good valuation and purchasing decisions.
One approach under consideration is that of a public-private partnership in a large “aggregator” bank (the “bad bank”), funded with both public and private money. Because private managers will run it, the bad bank could arguably produce a lower risk of overpayment than would direct purchases by the government.
The main problem with an aggregator bank is that it adds only one additional buyer, albeit a big one, to the market. Suppose that, due to the current lack of buyers, banks can sell troubled assets of a certain kind at a price—say 20 cents on the dollar—substantially below fundamental economic value. To the extent that the aggregator bank is run in a profit-maximizing way, the bank will push for a price as close as possible to 20 cents on the dollar; in this case, introducing the aggregate bank would not close the gap between prices and fundamental values as effective competition on the buyers’ side would be expected to do. Alternatively, if the aggregator bank is structured so that it does not seek to drive as hard a bargain as possible, but rather pays the “right price,” we are back to the problems of arbitrary valuation outside a market context: the very problems that rightly led to abandonment of the Bush administration’s plan.
Instead, plan for restart the market for troubled assets should be based on establishing a significant number of competing funds. Each of these funds should be privately managed and financed with both public and private capita, and the funds should have sufficient aggregate capital.
Suppose the government wishes to introduce an additional $500 billion of public and private capital into the buying side of the market for troubled assets. Rather than establish a single aggregator bank with $500 billion, the government can establish a significant number, say 25, of “bad bank” funds, each bankrolled with $20 billion in public and private funding. And suppose that each of these funds is run by a private manager that will capture a share of the profits generated by the fund (above the yield on Treasury securities) and possibly also bear a (lower) share of any losses.
The existence of such a significant number of private buyers armed with substantial capital will produce a market for troubled assets with many potential sellers (banks) facing a significant number of potential buyers (the funds). And the profit share captured by the funds’ private managers will provide these managers with powerful incentive to avoid overpaying for troubled assets.
At the same time, the profit motive of the selling banks, coupled with the presence of competition among the private funds, will make it difficult for funds to underpay for troubled assets. As a result, we can expect a well-functioning market, with prices set around the fundamental economic value of purchased troubled assets.
The question that remains is how to induce the participation of sufficient private capital in competing funds dedicated to buying troubled assets. Because private capital has not yet flown in large amounts into the market for troubled assets, the Treasury is contemplating arrangements to encourage such participation. This can be done by having the participating public capital assume more downside risk or, alternatively, capture less of the upside. And the critical issue is how to do so to the minimum extent necessary to obtain the desired capital.
To ensure that the government does not overspend, private managers should not only compete for troubled assets after they obtain capital from the government but also compete upfront for the right to participate and receive funding from the government’s program. Such a market mechanism can ensure that the government provides funding at a level and under terms that will be least costly to taxpayers while still inducing the establishment of private funds with the desired amount of aggregate capital.
Consider a simple scheme whereby the government’s program provides capital to new funds in the form of debt financing. Under this scheme, the program finances a specified fraction of the capital of each participating fund with a non-recourse loan. The terms of this loan will be similar to those used by the Federal Reserve in the facilities it recently established for funding pools of consumer loans: the loan will be paid first from the payoffs of the private fund; will be paid only from those payoffs (being non-recourse); and will carry a low interest rate.
To understand how such debt financing can induce private capital to invest in the private funds participating in the program, imagine that the government sets a very high level of 95% for the fraction of the participating funds’ capital to be provided in such debt financing. With such large government participation, it will presumably be easy to get the private side of each fund – the private manager and the private investors partnering with the manager – to contribute the remaining 5% of the fund’s capital as equity investment. In this case, while the private side will be the first to bear any losses of the portfolio, that potential loss will be capped at 5% of the fund’s capital, and the private side will fully capture any profits above the loan’s low interest that are generated by the 95% of the fund’s capital provided by the government’s program.
While providing 95% of a participating fund’s capital in the form of governmental debt financing is thus likely to be more than sufficient to attract private capital to the fund, doing so would likely impose on taxpayers a higher expected cost than is necessary. How then should the participation rate – the fraction of participating funds’ capital that the government will fund with debt financing – be determined? It should be set through a competitive bidding process.
Suppose that, while the government wishes ultimately to have $1 trillion in purchasing power in the hands of funds dedicated to purchasing troubled assets, it will begin with a “pilot” round in which private funds with an aggregate purchasing power of $100 billion are established. The government should invite bids from private managers seeking to participate in this round. Each bid should indicate, first, the maximum fraction of the fund’s capital that the private side commits to contribute as private equity capital (rather than using debt financing from the Investment Fund for this purpose), and, second, the size of the fund the private side seeks to establish.
Because it is desirable to have a significant number of different funds in the marketplace, any given private manager seeking to establish a fund will be able to get no more than a specified fraction, say 10%, of the aggregate capital provided to the initial round of funds. The government can also set, at least in the initial pilot round, a minimum level of equity capital contribution, say 10% of capital, below which bids for establishing a fund may not be submitted.
Once the bids are made, the government will set the level of its participation under the program at the lowest level that can be set while still allowing for establishing funds that collectively have the total target capital for the program’s initial round. Thus, for example, the government will set the equity contribution percentage at 40% and the government’s debt financing at 60% if, given the received bids, the 40% level, but no higher level, will result in establishing private funds that collectively have the target level of aggregate capital.
While the discussion thus far has assumed that the government will contribute only debt financing, and the private side will supply only equity financing to participating funds, the proposed scheme could allow for equity participation by the government or debt participation by the private side. Whatever combination is chosen, it will be important to have the key parameter that determines the government’s expected costs (such as the fraction of capital provided by the private side) set through a competitive process; this will keep the government’s expected costs at a minimum.
Another option to consider is dividing the government’s program into sub-programs dedicated to particular subsets of the universe of troubled assets; for example, rather than trying to get $500 billion to go after troubled assets in general, the government might seek to get $200 billion to go after CDOs, $100 billion to go after troubled mortgage assets, and so forth. In such a case, each of the sub-programs will use the proposed competitive mechanism for selecting the funds that it will support.
By inducing private managers to compete at two levels – first, to participate in the government’s funding program, and subsequently to buy banks’ troubled assets – the design proposed above can deliver the results the Treasury has announced it seeks. It will effectively restart the market for troubled assets and do so at the least cost to taxpayers.
The Discussion Paper that puts forward in detail the plan described above is available here.