It is not often you get pieces as well written and insightful as this one. Stuart Mackenzie worked at the NAB for three years and has 20 years experience in banking before he became a freelance journalist and this magnificent piece chronicles the NAB’s $3 billion Homeside debacle.
This is not the first time NAB has had difficulties in the US.
While analysts and the media were praising the bank for avoiding the worst of the domestic lending excesses of the 1980s, the NAB had built the largest US branch network of any of the Australian banks. In 1989, US assets totalled $4 billion, or 5 per cent of the balance sheet, but the operation reported a before tax loss of $12 million that year.
Over the next two years, assets were run-off to 1.5 per cent of the balance and branch offices in Dallas, Houston, Atlanta, San Francisco and Chicago were closed, but not without further losses – $24 million in 1990 and $14 million in 1991.
NAB chief executive Frank Cicutto was executive vice president Americas in 1988; executive general manager Bob Prowse was responsible for US operations in 1990; and Don Argus took over the top job from Nobby Clarke in October 1990.
As he watched the operation cut back to one branch in New York and a representative office in Los Angeles, senior US operations vice-president, Neal Popkin, took consolation from Argus’ promise that one day he would buy him a real bank to manage.
It took five years to deliver on that promise, with the $2.1 billion purchase of Michigan National Bank in November 1995.
Argus had little international banking experience when he was appointed executive banking director, and heir apparent to Nobby Clarke, in 1989. He participated in the due diligence on Clydesdale Bank’s loan book, before NAB took the number three Scottish bank off Midland Bank’s hands in the mid-1980s, but had not held a senior expatriate post in any of the NAB overseas operations.
If he was going to continue NAB’s international expansion strategy, Argus had to learn about banking in NAB target overseas markets the UK and the USA.
One of his US banking mentors was Hugh McColl of NationsBank.
McColl’s ancestors had fled to the Carolinas after the defeat of Bonnie Prince Charlie and the collapse of the Jacobite uprising at Culloden in 1746.
In his 18 years at the helm of NationsBank previously NCNB North Carolina National Bank McColl’s highland tenacity created one of the most acquisitive banks in US history.
In 1983 NCNB operated in two states, had US$12 billion in assets, and net operating income of US$92 million. By the time McColl retired in April this year, he had acquired dozens of banks, including the venerable Bank of America, in whose name the group now operates. Assets grew 50 times to US$610 billion, generating net operating income of US$7.9 billion from operations in 22 US states and 37 other countries.
With McColl as mentor, the boy from Bundaberg turned NAB into Australia’s most acquisitive bank.
The Michigan National Bank deal followed the successful purchase of the Bank of New Zealand three years earlier.
Both targets were commercial banks, whose business was well understood by NAB’s predominately retail banking executives.
Rather than trying to pick up bargains, NAB relied on its easy access to capital and a thorough due diligence process to buy the income stream of the acquisition and drive out synergistic value, largely through cost rationalisation.
“Analysts don’t understand our acquisition strategy, ” Argus said in a 1994 interview with Business Review Weekly.
“One day they will wake up and look at the franchises we have purchased, and at what price, and see the income we are driving out of them. They should look at the historic efficiency ratios and how far the National can drive those ratios down.”
In the same interview, Argus viewed technology as fundamental to his bank’s success.
“I recall, 12 months ago, sitting with some US bankers around a table, and they were talking about , 05cost-cutting and making an extra bit with their margins. None of this makes sense unless you have a capable technology platform underneath.”
Mortgage-processing technology was a major justification when the HomeSide deal was announced in October 1997.
Argus claimed that the acquisition would enable NAB to transform the Australian mortgage market and achieve major cost savings by integrating the group’s mortgage businesses around the world.
“We are purchasing a well-established business and a very professional management team,” he said. Most analysts and business media commentators bought the line.
“Of all the overseas acquisitions that National Australia Bank has made, the $1.7 billion deal announced yesterday makes, on paper at least, the most sense,” Stephen Bartholomeusz said in The Age.
Robert Gottliebsen in Business Review Weekly did ask why the Americans allowed a little-known Australian bank to buy this leading technology at a small premium and concluded, “The obvious answer is that they were not prepared to take the risk on unproved technology.”
But how did a little-known Australian bank find out about this jewel in the first place?
The ‘well established business’ that Argus had in his sights had been formed just 18 months earlier by merging the mortgage servicing businesses of BankBoston and Florida-based Barnett Banks. HomeSide was floated in January 1997, but only 19 per cent of stock was offered to the public. BankBoston and Barnett each retained 27 per cent and two US investment firms owned the remaining 27 per cent.
Meanwhile, Hugh McColl was in town and NationsBank’s target was Barnett Banks.
On 29 August 1997 McColl announced the (then) largest bank merger in US history by offering US$15.5 billion for Barnett, which would make NationsBank the third largest banking company in the country.
There was a plum job for Barnett’s chairman Charles E. Rice as chairman of NationsBank, while McColl remained chief executive of the combined banks.
The offer valued each Barnett share at around US$75 and their market price rose 32 per cent on the announcement.
But the deal was also good news for another Jacksonville-based stock.
HomeSide Inc.’s stock rose from under US$21 the day before the deal to a high of US$26 in the following week.
NationsBank-Barnett’s 27 percent stake in HomeSide was suddenly worth a lot more if a buyer could be found.
Local analysts speculated that NationsBank would seek to take over the entire company because of the complex contractual relationship between Barnett and HomeSide.
HomeSide serviced all mortgage loans originated by Barnett and roughly half of the loans under management were those originally transferred from Barnett and BankBoston to create HomeSide. After 3 years, Boston Bank and Barnett Banks had the option of taking their administration to another company.
NationsBank already owned a mortgage processor, however, and the opportunity to offload Barnett’s stake in HomeSide, at a recently inflated share price, to a willing buyer interested in acquiring the mortgage processing technology must have been attractive.
Argus announced the NAB’s A$1.7 billion cash offer of almost US$28 per share just two months after the Barnett-Barnett deal when the shares had been trading at under $21.
When NAB took them out, HomeSide’s original investors had made gains of almost 90 per cent in nine months.
Australian analysts and business media examined the deal.
Gottliebsen in BRW cautioned, “Obviously, Argus is taking a risk with technology that is not yet fully implemented in the business he is buying. If something goes wrong, it will affect the entire HomeSide business. Few other Australian boards and managing directors would be prepared to take such a risk.”
In the Australian Financial Review, Ivor Ries pointed out that $1.1 billion of the $1.7 billion price tag was pure goodwill and also identified the risk that in three years almost half of HomeSide’s customer base could take their business away.
But (then) Prudential-Bache Securities banking analyst Anna Borzi appeared to be a lone voice identifying the risk of buying into a securitised mortgage servicing business that was “new to the Australian bank sector.”
Mortgage servicing rights (MSRs) was not a term on everybody’s lips until Frank Cicutto revealed the first $870 million of losses in July this year.
By the time Cicutto announced a further $3 billion write down in September, everyone was talking about them.
NAB prides itself on the standard of disclosure in its annual reports. In 1998, 1999 and 2000, these provide figures on the value of HomeSide’s mortgage servicing rights, but the only relevant descriptive statement concerns accounting policy and includes the following:
“Mortgage servicing rights , 05 are evaluated for impairment by comparing the carrying amount of the servicing rights to their fair value. Fair value is estimated using market prices of similar mortgage servicing assets and discounted future net cash flows considering market prepayment rates, historic prepayment rates, portfolio characteristics, interest rates and other economic factors. , 05The mortgage servicing rights are stratified by the predominant risk characteristics [and] impairment is recognised through a valuation reserve for each impaired stratum and is included in amortisation of mortgage servicing rights.”
Compare this with the following extract from HomeSide’s form 10-K filing with the US Securities and Exchange Commission for 1998:
“, 05the value of the , 05 mortgage servicing rights tends to diminish in periods of declining interest rates and increase in value in periods of rising rates. This tendency subjects HomeSide to substantial interest rate risk. It also directly affects the volatility of reported earnings (our emphasis), 05It is HomeSide’s policy to mitigate and hedge this risk through its risk management program.
“The risk management instruments used, 05have characteristics such that they tend to increase in value as interest rates decline. Conversely, [they] tend to decline in value as interest rates rise. Accordingly, changes in the value of these hedge instruments will tend to move inversely with changes in value of HomeSide’s mortgage servicing rights.”
Clear and unambiguous. There is a substantial interest rate risk in MSRs. They go down in value when interest rates decline and up in value when rates increase. The risk management contracts used by HomeSide to hedge this exposure do the reverse, so hedge profits should be expected during periods of declining rates and hedge losses expected when rates are rising.
NAB’s shareholders, not to mention the analysts and the media, may have preferred this explanation to the arcane statement they got. Key NAB executives were aware of it, however Don Argus, Bob Prowse and Frank Cicutto were directors of HomeSide from 1998 and NAB chief financial officer Richard McKinnon joined the board in 2000.
NAB shareholders had to wait until $4 billion of their money had been lost, before the risk management section of the annual report contained any qualitative description of mortgage servicing rights risk and the difficulties of its hedging.
Furthermore, to find even the incomplete information, shareholders had to read the detailed annual report. Since 1999, NAB has given shareholders the option of receiving a concise annual report one that, according to the bank’s shareholder centre, most individual shareholders take up.
The board of directors as a whole is responsible for the content of both the concise and full annual report and the principal executive engaged on its preparation is the chief financial officer – Bob Prowse in 1998 and 1999, then Richard McKinnon for the 2000 and 2001 reports.
KPMG gave the full accounts an unqualified audit report for all years and reported to shareholders that the concise report was ‘free from material misstatement.’ KPMG also audited HomeSide’s accounts from April 1998.
The 2000 annual report, with no explanation about MSR risk, was signed on behalf of KPMG by the same Christopher Lewis who became NAB’s executive general manager, group risk management in July this year.
The trail of knowledge about the NAB’s exposure to the ‘substantial interest rate risk’ since 1998 can be traced from the HomeSide board through key members of NAB’s executive team to the main board, but not on to NAB shareholders. Auditor KPMG’s knowledge should have come from their examination of both HomeSide’s and the NAB group accounts over the same period, which includes both Argus’ and Cicutto’s stewardship.
If the risks were known, how should they have been monitored and managed? Every time a bank takes a deposit, writes a loan, or buys or sells foreign exchange and securities, it assumes risk from its customers. Managing this financial risk is a core business for a bank.
Only a few of the risks to which a bank is exposed are regularly discussed in the media. Credit is the most obvious one the risk that a borrower will default on a loan repayment. Recent major corporate collapses such as HIH, Ansett and Enron have again bought this into focus and questions are being asked about bank’s exposures to those companies and the extent of their bad debt provisioning.
In the early 1990s, there was much excitement in the media about the trading of derivative instruments, prompted by a few spectacular losses by both banks and their customers.
However, a far greater risk to a bank’s performance, or even survival, is rarely discussed. It is usually referred to as structural interest rate, or balance sheet risk.
While the amounts at risk from defaulting borrowers, even large corporates, may be in the order of hundreds of millions of dollars, the stakes in balance sheet risk are in the billions.
Every deposit and loan that the bank makes creates an exposure to movements in interest rates. Loans and deposits are made at fixed or variable rates and for fixed periods or at a call. While individual banks determine the interest rates they will offer to customers, these are based on rates in the professional money markets, with a bank-determined margin added.
When all of a bank’s assets and liabilities are aggregated together, many of the individual exposures will net each other out, but ones that do not form one large risk position. A bank manages this risk by executing transactions in the professional money markets, in just the same way that a money market dealer would manage a trading position – borrowing and lending in the inter-bank money markets; entering into derivative contracts such as futures, options or swaps.
So after all the product marketing, retail lending and deposit taking are completed through the branch network, the bank’s success or failure depends on a few risk management and financial markets experts.
The positions and risks involved are huge and their management is not entrusted to any one person in the bank. In most banks the responsibility falls to a committee referred to as the asset and liability committee and commonly abbreviated to ALCO.
ALCO frequently comprises a subset of the bank’s executive committee those responsible for managing the lines of business supplemented by economists and market risk experts.
A small department, the asset and liability or balance sheet management unit (ALM), supports ALCO. They aggregate data from the bank’s main computer systems a significant task in itself in a large bank into specialised computer models that allow the effects of an overall rise or fall in rates, or a change in the relationship between short and long-term rates, on the bank’s profit and loss to be predicted.
In the light of these forecasts, ALM will provide options for managing the balance sheet risk to ALCO for decision-making.
ALM will implement those decisions by executing transactions in the financial markets either directly or by placing orders with the bank’s treasury dealers.
The effectiveness of those transactions will be regularly monitored by ALM, reported on at the next ALCO meeting, and so the cycle continues measure the risk, take corrective action, monitor the results.
The people in the ALM function are some of the most highly qualified staff in the bank. Some are accountants, but not many because the mathematics involved is beyond the experience of most bean counters and most bankers. They are adept at advanced mathematics many hold higher degrees and actuarial qualifications. Most have never worked in the retail part of the bank their career path is more likely to have been from a graduate intake programme into the bank’s treasury or finance divisions. They live and breathe the same world of financial markets as the treasury dealers the ‘phone jockeys’ berated by the media and politicians for unpatriotically selling the Australian dollar to make a profit.
And they hold the performance of the bank in their hands. One slip, one data entry error into a model, one wrong interest rate assumption and a bank can make or lose billions. It is management’s responsibility to ensure that an effective system of checks and balances is in place.
Well managed, ALM can make huge profits for the bank. Former deputy treasurer at the State Bank of NSW, Graham Hand, provides an inside account of how that bank’s ALM kept the bank afloat during the lending crises of the 1980s in his recent book Naked Among Cannibals.
And at NAB?
The bank has an established process for managing acquisitions, although executive responsibility for it varied during the 1990s. It was part of group strategic development in 1993, was transferred to group finance after the arrival of new chief financial officer, Bruce McComish, in 1994 and back to strategic development under group general manager Roland Matrenza in 1997.
Current CFO, Richard McKinnon, was NAB’s point man for the due diligence when he was head of investments and advisory. The risk expertise on the HomeSide team included the (then) head of ALM for the NAB’s domestic operations, John Holihan, who is now head of internal audit for wholesale banking.
McKinnon was also the NAB’s key executive, during Argus’ watch, on another ill-fated investment in financial technology its joint venture with John Maconochie. McKinnon, Argus and Cicutto are all named defendents in a $56 billion court case that may collapse because Maconochie has run out of money, but which is reported to have cost NAB $55 million in legal costs to date.
But back to HomeSide.
With the purchase completed, the ongoing management of the balance sheet risk should have proceeded along the lines described above.
MSR’s are an asset on HomeSide’s balance sheet and, according to the bank’s annual report, each regional bank manages its own structural balance sheet risk under policies and limits set by the group asset and liability committee (ALCO), under delegated authority from the Board.
Head office overview of the group’s global structural balance sheet risk is managed by a group balance sheet management unit and reported on a “regular basis” to ALCO and “on a periodic basis” to the board.
The group balance sheet management unit has been part of group finance since 1994, reporting indirectly to the chief financial officer Bruce McComish until early 1998, Bob Prowse from 1998-90 and Richard McKinnon since then.
Chris Matten headed up group capital and balance sheet management from November 1997 until his resignation late last year. Matten has risk management experience with Japanese and Swiss banks and is the author of a respected text on managing bank capital. He was seen as a future CFO, but took up a similar role at Singapore’s OCBC Bank in January, after McKinnon got the NAB job.
Matten’s group should have been oversighting HomeSide’s own ALCO which was accountable to then chief executive Joe Pickett.
Pickett’s reporting line was directly to the group chief executive Argus until June 1999 and then Cicutto.
Oversight in this context should encompass several activities, and in 1998 the Australian Prudential Regulation Authority (APRA) issued useful guidelines for the use of internal risk models.
These include ensuring that: the processes and models are conceptually sound; the bank has sufficient staff in the risk control, audit and back-office areas who are skilled in the use of sophisticated models; the models have a proven track record; and they are regularly stress tested.
Comparing a model’s operation to another, perhaps industry standard, model whose effectiveness is already understood is a common means of testing its conceptual soundness, and is referred to as calibrating the model.
Stress testing should be conducted to identify events or influences that could greatly impact the bank. APRA recommends that:
“Banks’ stress scenarios should cover a range of factors that can create extraordinary losses or gains in , 05 portfolios, or make the control of risk in those portfolios very difficult.” These should include “, 05low-probability events including, 05past periods of significant disturbance (for example, the 1987 equity crash, or the fall in bond markets in the first quarter of 1994) and incorporate both the large price movements and the sharp reduction in liquidity associated with these events.”
At a more mundane level, basic auditable internal controls should be in place to prevent, or rapidly identify, data entry errors.
APRA emphasises the importance of the risk measurement system being transparent and accessible. Those in oversight roles, including auditors, should be in “, 05a position to have easy access, whenever they judge it necessary, to the model’s specifications and parameters.”
The objective of these controls is to ensure that the risk models used do what they are intended to, that their limitations in periods of extreme or volatile market conditions are understood, and that appropriate check and balances exist over the management of a critical area of risk exposure.
NAB’s explanations to date for the $4 billion loss at HomeSide include: data entry errors; input of wrong assumptions; a modelling error discovered by external consultants; a breakdown of the hedging model under extreme market conditions.
If effective, industry standard internal controls were in place and followed at group level in Melbourne none of these events should have occurred.
NAB’s 2001 annual report confirms that HomeSide’s MSR risks cannot be fully covered by any hedging strategy, and are ‘actively managed’ under balance sheet management policies approved by the board and monitored by the group assets and liability management committee (ALCO).
Additionally, group risk management prepare an annual ‘group risk profile’ for the board on major risks and procedures for their day-to-day management, control and responsibility.
The board is responsible for having an “in-depth understanding of each substantial area of the business,” and for ensuring adequate risk management processes are followed, according to the NAB’s corporate governance statement.
Any individual director, who is unfamiliar with the nature of the risks concerned, can seek the chairman’s approval to obtain independent external advice under NAB’s board guidelines.
The trail of knowledge through NAB’s executive committee via the chief executive to the board not only includes the nature of the MSR risks, but regular reports of the exposure incurred and the procedures and controls for managing the risk.
Three HomeSide executives have paid the price for the management failures in Jacksonville. The nightmare on the executive and boardroom floors of NAB’s Bourke Street head office may drag on, as it appears that the chairman’s report will not be ready in time for Thursday’s annual general meeting.
Clarification: Stuart Mackenzie’s piece on the NAB’s HomeSide problems stated that Tsu Ming Lao was a member of the bank’s pre-acquisition due diligence team. We understand that this was not the case, and Mr Lao’s involvement was not until May of this year, as part of the internal investigation team after the hedging problems at HomeSide had been discovered. Stuart and Crikey apologise unreservedly for the error and the story has been appropriately corrected.
* Stuart Mackenzie is a freelance journalist. He has over twenty years of international bank management experience in the UK, the Channel Islands and Australia. He is a Fellow of the Institute of Chartered Accountants in England and Wales and an Associate Fellow of the Australian Institute of Management. He was employed by NAB’s Group Treasury from 1993 to 1996 and currently holds shares in the bank. He can be reached at [email protected]