Northern Rock

A building society turned growth stock

Northern Rock began as a mutually-owned building society, formed in 1965 from the merger of two older Victorian societies. It demutualised and listed on the London Stock Exchange in 1997, and grew rapidly from there. Between June 1998 and June 2007 its total assets grew from about £17.4 billion to about £113.5 billion, roughly 6.5 times larger, an implied compound annual growth rate of about 23%. Retail deposits, the ordinary savings accounts that fund most banks, grew far more slowly over the same period, from about £10.4 billion to about £24 billion. The share of the balance sheet funded by stable retail deposits fell from about 60% in 1998 to about 23% by June 2007. Northern Rock filled the rest of the gap with securitised mortgage notes and, increasingly, short-term wholesale borrowing that had to be renewed every few months.

A trigger from outside the bank

The immediate trigger arrived from outside Northern Rock altogether. On 9 August 2007, BNP Paribas suspended three investment funds because it could no longer reliably value their exposure to US subprime mortgage assets. Interbank and short-term funding markets across Europe seized up almost immediately. Within days, Northern Rock's managers told the Financial Services Authority (FSA) that they were having funding difficulties, and the Bank of England was informed the next day. The bank's mortgage book itself was not the underlying problem: its assets were, on the regulator's own later assessment, good quality. The problem was that Northern Rock could no longer roll over the short-term wholesale borrowing that funded a large share of those assets.

The announcement that triggered the run

On the evening of 13 September 2007, the BBC reported that Northern Rock had sought Bank of England support. The next day, the Bank of England, HM Treasury and the FSA jointly confirmed a liquidity support facility, describing it as a lender-of-last-resort operation, against appropriate collateral and at a penalty interest rate, and stating that the FSA judged Northern Rock solvent, above its capital requirement, with a good quality loan book. The announcement was intended to reassure depositors; it had the opposite effect. Savers queued outside branches from 14 to 17 September, Britain's first bank run in a century. An estimated £1 billion was withdrawn on the first day alone, with further billions following over the weekend. On 17 September, Chancellor Alistair Darling announced that the government would guarantee all existing Northern Rock deposits, which finally stopped the queues.

From guarantee to nationalisation

The damage to the funding side of the balance sheet did not stop there. By the end of December 2007, wholesale funding had fallen from £26.7 billion to £11.5 billion, a drop of about 57%, and retail deposits fell by almost exactly the same proportion, from about £24.4 billion to about £10.5 billion. A Bank of England loan plugged that gap, standing at £28.5 billion by the end of the year. The authorities explored private-sector rescue bids but judged none of them gave enough assurance for the public money already committed, so on 17 February 2008 the Chancellor announced that Northern Rock would be taken into temporary public ownership. The formal transfer took effect at 00:01 GMT on 22 February 2008, the first UK bank nationalisation since the 1970s. Shareholders fared worst of all: the Treasury's appointed valuer later set the compensation value of Northern Rock's shares at nil, and shareholder legal challenges were rejected on appeal.

No option, future or swap sits at the centre of this case, and that absence is exactly why it belongs in the course. Northern Rock is the clearest real picture of the liquidity-preference and duration-mismatch mechanics behind the yield curve, and of the same margin-turns-into-liquidity theme that runs through the wider 2007-08 credit crisis.

Yield curves usually slope upward for reasons that include liquidity preference: savers prefer to lend short, so they can get their money back quickly, while borrowers prefer to borrow long, so they are not forced to refinance often. A bank normally sits between the two and manages that mismatch carefully. Northern Rock took it to an extreme. It wrote long, fixed-term mortgages, an asset with a long effective duration, and funded a shrinking share of that book with stable retail deposits, filling the rest with securitised notes and short-term wholesale borrowing that had to be rolled over every few months. By June 2007, only about 23% of its liabilities were ordinary retail deposits.

The distinction that matters is that this is a funding, or rollover, version of a duration mismatch, not a market-risk one. Northern Rock's assets were, by the regulator's own account, solvent and good quality. The failure was that the short-term liabilities funding those long assets could not be renewed. When wholesale and securitisation markets froze in August 2007, Northern Rock could not roll its short-term borrowing, and a system-wide credit event became, for this one bank, a liquidity event. A solvent bank ran out of cash.

The queues outside branches are the image most people remember, but they were a second-order event. The 14 September announcement of Bank of England support caused the public run, not the other way round: the wholesale funding problem had already been building for a month before the queues formed. Once collateral and funding lines could no longer be relied on, a credit problem became a liquidity problem, the same mechanism that runs through the rest of the 2007-08 crisis. Northern Rock is the UK-specific instance of it, working through mortgage-backed wholesale funding rather than repo haircuts, but the lesson is the same: margin and rollover risk convert a credit problem into a liquidity problem.