Global liquidity provision is highly procyclical. The recent financial crisis has resulted in a flight to safety, with severe strains in key funding markets leading central banks to employ highly unconventional policies to avoid a systemic meltdown. Bagehot’s advice to “lend freely at high rates against good collateral” has been stretched to the limit in order to meet the liquidity needs of dysfunctional financial markets. As the eligibility criteria for central bank borrowing have been tweaked, it is legitimate to ask, How elastic should the supply of central bank currency be?Even when the central bank has the ability to create abundant official liquidity, there should be some limits to its support for the financial sector. Traditionally, the misuse of the fiat money privilege has been limited by self-imposed rules that central bank loans must be fully backed by gold or collateralized in some other way. But since the onset of the crisis, we have seen how this constraint has been relaxed to accommodate the demand for market support. My suggestion is that there has to be some upper limit, and that we should work hard to find guidelines and policies that can limit the need for central bank liquidity support in future crises. In this paper, I review the recent expansion of central bank liquidity support during the crisis, before discussing the collateral polices related to central banks’ lender-of-last-resort and market-maker-of-last-resort policies and their rationale. I then examine the relationship between the central bank and the treasury, and the potential threat to central bank independence if they venture into too much risky balance sheet expansion. A discussion about the exceptional growth of the shadow banking system follows. I introduce the concept of “liquidity illusion” to describe the fragility upon which much of the sector is based, and note that market growth has been based largely on a “fair-weather” view that central banks will support the market on rainy days.I argue that we need a better theoretical framework to understand the growth in the shadow banking system and the role of central banks in providing liquidity in a crisis.Recently, the concept of “endogenous finance” has been used to explain the strong procyclical tendencies of the global financial system. I show that this concept was central to Hyman P. Minsky’s theory of financial instability, and suggest that his insights should be integrated into the ongoing search for a better theoretical framework for understanding the growth of the shadow banking system and how we can limit official liquidity support for this system. I end the paper with a summary and a discussion of some of the policy issues. I note that the Basel III “package” will hopefully reduce the need for central bank liquidity support in the future, but suggest that further structural reforms of the financial sector are needed to ease the tension between freewheeling private credit expansion and the limited ability or willingness of central banks to provide unlimited official liquidity support in a future crisis.