Tuesday I began with a simple example, which I am calling Large Bank.

Yesterday we looked at TARP I, in which the government would buy troubled/toxic assets from banks.

Today I will describe TARP II, the plan we (the Bush Administration) implemented, in which the government made direct equity investments in banks to help fill their capital holes. We called this the Capital Purchase Program.

As a reminder,we are trying to address two problems:

  1. Large Bank does not have enough capital.
  2. Large Bank has downside risk on its balance sheet due to the uncertain value of these bad loans. That downside risk makes the firm’s value uncertain and scares away investors.

Here is the balance sheet for Large Bank:

Assets Liabilities and Equity
Good loans 800 Deposits 600
Bad loans 120 Debt 300
Equity 20
Preferred 0
Total 920 . . . . . . . Total 920
  • Total capital: 20
  • Leverage: 46:1

TARP I solves problem #2 if you buy all the bad loans, but it is extremely inefficient in addressing problem #1, the capital hole. Spending 120 from the TARP to buy the bad loans would provide no new equity capital. Spending 150 to buy the loans valued at 120 would provide a net 30 of capital for 150 outlayed from the TARP.

The constraint is not the ultimate cost to the taxpayer. It is instead the legislated limit on how much outstanding cash can be invested/spent at any one time from TARP: $700 B.

To fill the capital hole, our first choice would be for the bank to attract private capital. Bank management appears reluctant to do this, because they don’t want to dilute the value of their existing shareholders. In addition, private investors were unwilling (at least last Fall) to invest in banks that had significant downside risk on their balance sheets. So temporary public capital, provided by the taxpayers, was the only option to recapitalize the banking system.

If we take the same 120 from the first TARP I case, but instead use it to buy preferred stock in Large Bank, we end up with this:

Assets Liabilities and Equity
Good loans 800 Deposits 600
Bad loans 120 Debt 300
Cash 120 Equity 20
Preferred 120
Total 1,040 Total 1,040
  • Total capital: 140
  • Capital added by this transaction: 120
  • Leverage: 7.4:1
  • Risk of bad loans: still lies entirely with Large Bank
  • Taxpayer outlay from TARP: 120
  • Long-term cost to taxpayer: Zero if the firm remains solvent, up to 120 if the firm goes bankrupt.


  1. Large Bank has 120 more capital from the government. It is now well capitalized and has a good leverage ratio. (I am for now glossing over the difference between preferred and common stock.)
  2. Large Bank still has all the downside risk associated with the bad loans. This may continue to scare away private capital, depending on estimates of the relative size of the strengthened capital cushion and the downside risk of those bad loans.
  3. The government has spent 120 of the TARP pool.
  4. The taxpayer will get dividends from the preferred stock (which look a lot like interest payments at a fixed interest rate).
  5. The government is now the majority investor in Large Bank and has both an ongoing taxpayer interest in and leverage over how the bank is run.

TARP II gets tremendous bang for each TARP buck in recapitalizing banks. If you are more worried about the capital hole than the downside risk, then TARP II has far better arithmetic.

If you are worried that the capital problem being bigger than you think, then you want to use TARP resources in the most efficient way possible. Remember that you do not have good information about the size of either the capital hole or the downside risk. You know what the banks report about their balance sheet, but especially last fall, we had to be extremely skeptical about the information being reported about the size of both problems. This is one reason why the stress tests are so valuable — regulators and policymakers presumably have much better information now about the absolute and relative sizes of each problem, at least for the 19 largest banks.

Over the past few months we have been experiencing a significant policy downside of direct equity investment, and it is a major difference between TARP I and TARP II. Under TARP I, the government buys the asset and the relationship between the government and the bank ends. Under TARP II, the government has an ongoing relationship with the bank. This creates policy tension among three different governmental roles:

  1. government as rule-setter and regulator;
  2. government as investor on behalf of the taxpayer; and
  3. government as an interested party with other policy (or non-policy) goals.

This tension is playing out in several uncomfortable and unpleasant ways. The government is involved in compensation decisions within the firm. The government is leveraging its investment to pursue other goals, like encouraging the banks to lend and maybe leveraging some of them to write down auto manufacturers’ debt. And the government may trade off other policy goals against the taxpayer’s investment by allowing banks to convert preferred equity to common equity.

Returning to the original two goals, you can see the two extremes in TARP I and TARP II. TARP I was focused on removing downside risk from balance sheets. The danger with TARP I is you might buy $700 B of bad assets, and have only made a medium-sized dent in the downside risk problem, while doing far less to fill the capital holes. TARP II ignores the downside risk problem while getting a huge bang in filling capital holes. TARP II also has other problems derived from the ongoing linkage between Uncle Sam and the banks.

Tomorrow I will describe TARP III, the Geithner approach, and how it tries to address both problems by tapping into non-TARP resources.

As in the rest of this series, I thank Donald Marron for his examples and assistance.