"Banks Face Capital Constraints, Not Liquidity Constraints"
I have heard the above statement several times since the financial crisis began. The indefatigable von Pepe was ensconced at the library today and sent me several quotes from the work of Gerald O’Driscoll, including this:
The credit crunch continues because many banks are capital-impaired, not illiquid. Treasury’s policy shifts and inconsistencies under both administrations have sidelined potential private capital. Treasury became the capital provider of last resort. It was late to recognize the hole in banks’ balance sheets and consistently underestimated its size. The need to provide second– and even third-round capital injections proves that.
In another email von Pepe sent me a quote where O’Driscoll says that banks don’t “lend out” their capital, but rather the capital is a fund with which to absorb future losses.
In a previous post I went over the basics of recapitalization, but I thought it might be worth summarizing the discussion von Pepe and I had, just to make sure we agreed on exactly what O’Driscoll is talking about.
(A quick caveat: If you feel the need to say “duh!” in the comments, fine go ahead. With this stuff the issue was never that we doubted the truth of these claims, but rather that we wanted to just double check with each other to make sure we really understood them, and also to relate them to other concepts we had filed away. It’s a good safety measure to occasionally test whether you can reconcile one thing you “know” with other things you “know.”)
So anyway this might help some of you clarify your thoughts on this: Let’s say a bank is very liquid but undercapitalized. E.g. maybe it has $999 million in liabilities and $1 billion in assets (which include loans). So that means there is only $1 million in shareholder’s equity–“capital”–in the operation. If the bank’s assets fall in value by anything more than 0.1%, the shareholders are wiped out and the firm is insolvent.
In this scenario, having the ability to borrow large amounts of money at 0% interest doesn’t really do that much. E.g. let’s say the bank borrows $1 billion from the Fed at 0% interest, and uses it to buy one-year Treasurys yielding 0.5 percent interest. Then initially the bank’s liabilities go up to $1.999 billion, and its assets go up to $2 billion. The shareholder equity has been even further diluted, because the bank’s leverage has just doubled.
After the year passes, so long as everything else stays the same, the bank collects $1.005 billion from the Treasury and pays off the Fed loan. So now the bank’s liabilities are back down to $999 million and its assets are up to $1.005 billion, the difference being the small spread it earned borrowing from the Fed and lending to the Treasury for a year. But the bank is still way way undercapitalized.
In contrast, suppose that the initial bank doesn’t borrow money (from the Fed), but instead seeks outside investors to buy stock in the bank. So the bank takes in $1 billion in new capital by issuing new shares of stock. The moment after the new stock issuance, the bank has $999 million in liabilities and $2 billion in assets, $1 billion of the original assets (loans etc.) and the new $1 billion in cash. The shareholder’s equity has jumped from $1 million to $1.001 billion.
Now, the bank can go ahead and lend out that new $1 billion; it’s not that doing so will all of a sudden “drain away” the new capital they took pains to raise. After lending out the $1 billion in cash that the new investors pumped in, the banks liabilities are unchanged at $999 million, while the assets are still $2 billion. All that the lending has done is transform $1 billion of assets in cash, into $1 billion of assets in loans. The bank’s leverage is no problem at all in this scenario, and the bank would gladly lend out these funds, not just to the Treasury but even riskier projects if the potential return is high enough.
I think the real trick in all this is to realize that when a bank borrows money (from the Fed or anybody else), its liabilities go up because it has to pay back the loan. Thus, if a bank is trying to dig itself out of hole through earning the spread on interest rates, it will necessarily increase its leverage.
In contrast, if the bank raises new capital, its assets increase but its liabilities don’t. (The balance sheet still balances, of course, because remember that Assets = Liabilities + Shareholder’s Equity.) So the very act of raising new capital decreases the bank’s leverage. Then, when the bank lends out the newly-raised funds, this act doesn’t increase its leverage (though it reduces its liquidity). All that happens is that the composition of the assets changes.
I am writing this late at night and I have other work to do, so I reserve the right to fix anything dumb I have said above. It should go without saying that any errors are my fault and not von Pepe’s. But I hope the above discussion has helped some of you.
