Parliamentary Commission on Banking StandardsWritten evidence from Andy Hornby

I am writing to respond to the questions contained in the appendix to your letter dated 9 October 2012.

Section 1. Personal

I Joined Halifax in November 1999 and became Chief Executive of the HBOS Retail division in 2001 following the merger of Halifax and Bank of Scotland. I was Chief Executive of HBOS plc from August 2006 until the purchase by Lloyds.

It should be stressed that I have answered all the questions below to the best of my ability. I have not been provided with any papers to assist me in my responses (save that I have read the FSA Final Notice dated 9 March 2012) and many of the questions refer to a period of time many years ago hence my personal recollection of events is severely impacted by the passage of time.

Section 2. Growth of the Business

Questions 2(1) and (2).

The formation of HBOS plc in 2001 (created from the merger of Halifax and Bank of Scotland) was portrayed to the financial markets as the creation of a new force in banking. The concept was simple. Halifax was the largest UK provider of residential mortgages and liquid savings but lacked any significant presence in Corporate banking or SME banking. Bank of Scotland had significant Corporate banking expertise but lacked the deposit base and capital strength to fulfil its growth potential. The merger appeared to provide a marriage of Halifax’s capital strength and liquidity strength with Bank of Scotland’s Corporate banking credentials.

While the Group did pursue a growth strategy from its formation in 2001 to approximately 2005 to 2006 (with slower growth targeted from approximately 2005–6 onwards) the level of growth and shape of growth did vary by division. I shall answer in most detail on Retail banking as that was under my direct responsibility from 2001 to 2006 but also cover other divisions from 2006 onwards.

Following the creation of HBOS in 2001 the Retail division pursued a multi brand strategy for both deposit gathering and lending. In Retail banking there was as much emphasis placed on the growth of deposits (in order to reduce wholesale funding requirements) as there was on asset growth. The pace of growth in Retail banking was also adjusted to reflect the changing economy, with a lower share of the residential mortgage market targeted from approximately 2005 onwards. Greater restrictions were imposed in terms of limiting loan to value ratios as the economy showed signs of deterioration from approximately 2006 onwards.

The Corporate division was able to grow quickly because it was able to utilise the larger balance sheet capacity which resulted from the merger. With the benefit of hindsight it is clear (as witnessed from the extent of ensuing bad loan provisions) that the concentration of activity around commercial property, concentration in lower rated exposures, large single name borrower concentration, and exposure to highly leveraged transactions did mean that the division was vulnerable to a major downturn. Furthermore there was a further structural weakness in the European commercial property markets driven by the fact that most of the major providers of the lending to the sector were highly reliant on wholesale funding . When wholesale markets contracted sharply from 2007 onwards, the removal of liquidity from commercial property markets caused a further serious contraction in commercial property values. The combination of the sharply contracting wholesale funding markets and the Corporate division’s concentration in commercial property and lower grade assets was the primary cause of the ensuing provisions.

The Treasury division’s activities were concentrated around two core activities of providing funding and liquidity to the Group. From the time of the merger in 2001, HBOS had a considerable reliance on wholesale funding and this reliance on wholesale funding, in the event, proved to be the biggest vulnerability facing the Group from the onset of the financial crisis in 2007. However, the reliance on wholesale funding was recognised as a risk and actions were taken by the then Treasury and Finance Director to lengthen the longevity of our wholesale funding profile. More detail on the Treasury division’s activities is given in Section 5.

In relation to the International division, the lack of international presence (which led to HBOS being perceived as over-exposed to the UK economy) was perceived as a strategic weakness for the HBOS from the inception of the Group in 2001. From memory the UK provided well over 90% of profits in 2001 and a process of internationalisation (albeit from a very low base) began. Whilst asset growth overseas was far higher than asset growth in the UK, I estimate that International profits were still only around 10% of Group profits in 2008. More detail on our International strategy is given in Section 5.

Question 2(3)

The annual business planning cycle took several months. The divisions would begin by drawing up their own individual plans (including asset growth targets, risk parameters, and profit targets). All the divisional plans were then pulled together and discussed by the Executive Committee (a committee comprising the executive directors and the Group Risk Director). A highly detailed five year Business Plan was then produced (which described every divisional plan in depth) and this was presented to the Board.

The five-year Business Plan was presented to the Board in November each year (so for example the 2005–2009 business plan would be presented to the Board in November 2004). This document would include a detailed description of both asset and liability growth targets for every division and for the Group as a whole. The whole Board was therefore well involved in the creation of each five year business plan and the growth targets that lay behind the plan.

Furthermore during the course of the financial year every division would do at least one major set piece presentation to the HBOS Board whereby the growth strategy and risk appetite of the division was described in detail so the entire Board was well informed of the core strategies and risks of each division.

Question 2(4)

This question is difficult to answer without access to the relevant documents (particularly the regular feedback from shareholder visits which were carried out following the half year and full year results). I should also stress that I was only party to regular one to one interaction with shareholders from August 2006 when I became Group CEO. From memory, however, the investor relations audits of investor views generally produced very favourable responses from investors. The strong share price performance of HBOS versus its peer group between approximately 2001 and 2006 was testimony to the degree of shareholder support given to the key tenets of the HBOS strategy. From approximately 2007 onwards the shareholder interactions tended to be more concentrated around the Group’s response to the global financial crisis and, in particular, our funding and liquidity profile.

Question 2(5).

In general there was a broad degree of unanimity around the Boardroom table as to the company’s strategy (and its Business Plan which included the relevant growth targets). I have sat on a number of Boards as both an executive and non-executive director and I would view the HBOS Board as having been chaired in a broadly consensual style which allowed for executive and especially non-executive directors to challenge in whatever way was appropriate. Furthermore non-executive directors had further opportunities to challenge the strategies and risk appetites of individual divisions through the Risk Control Committee structure whereby every division had its own Risk Control Committee which was chaired by a non-executive director of HBOS and every divisional Risk Control Committee also contained at least one other non-executive director (see section 3 for more details). As an executive director I never felt that the non-executive directors felt restrained from challenging the executive team in whatever way was most appropriate.

Question 2(6).

In a business the size of HBOS, the communication to junior staff inevitably varied by division and sub-division within the company. I can speak best for the Retail division which I oversaw directly until 2006. Individual targets for deposit gathering, loan origination, customer service metrics and risk management measures would typically be held at an individual branch level and the branch manager would ensure that all branch staff understood the targets and the metrics. If a branch manager did not understand the reasoning for the targets/objectives which the branch had been set, then it would be the responsibility of the area manager to explain any inconsistencies. For clarity, it was the responsibility of divisional local management to communicate objectives through to junior level.

Section 3. Risk Management

Questions 3(1) and (2)

When the HBOS Group was created in 2001 it was based around a largely “devolved model” whereby each Divisional CEO was responsible for developing their respective processes and policies for managing risk at the divisional level, albeit within the framework of the Group’s overall strategic framework.

Each division contained divisional risk teams (stretching across regulatory, credit and operational risk) and the divisional risk teams were the first line of defence.

The Group Risk team (and the Group Credit Risk Committee in relation to credit risk) provided the first line of central challenge to the divisional executive and risk teams.

In order to provide a process of “external challenge” to the divisional risk strategies, from 2001 HBOS had a structure of divisional Risk Control Committees. Each division had its own Risk Control Committee which was chaired by an HBOS non-executive director and which contained at least one other HBOS non-executive director. In addition to the non-executive population, the divisional Risk Control Committees would have been attended by the most senior executives from the division and representatives from Group Risk and Group Internal Audit. The participation of the non-executives meant that the individual growth and risk management strategies of each division should be clearly understood by a number of non-executive directors at more detailed levels than would have been achievable from standard Board presentations. The concept was simple...for the non-executive directors to provide a check and balance to the executive members of the divisional Risk Control Committees. When the divisional Risk Control Committees met, this provided an opportunity for the non-executive directors who sat on the relevant Risk Control Committee to interrogate the rigour of the divisional risk strategies including credit risk, operational risk and regulatory risk issues.

Members of both Group Risk and Group Internal Audit would attend the divisional Risk Control Committees. So the non-executive directors also had an opportunity to hear the views of both Group Risk and Group Internal Audit as well as from divisional risk personnel. Finally it should be stressed that the Chairs of the individual Risk Control Committees (all of whom were non-executive directors of HBOS) reported findings directly to the Head of the Audit Committee as the Audit Committee had oversight of the divisional Risk Control Committee structure in order to ensure independence.

Inevitably the effectiveness of the various divisional Risk Control Committees was partly a function of the level and quality of information which was provided to them . I had most experience of the Retail Risk Control Committee (because as Retail CEO from 2001 I sat on that committee for many years). I felt that the Retail Risk Control Committee operated effectively. Its efficacy was helped hugely by the ready availability of data on retail markets (eg house price data going back decades). It was also helped by the relatively less technical nature of the risks inherent within Retail banking than within Corporate banking, Insurance and Investment or Treasury. It was therefore my perception that the non-executive directors on the Retail Risk Control Committee were clearly informed and understood the core risks within the Retail division.

The level of understanding in the non-executive director population regarding the level of risks inherent within some of the other divisions (particularly more technical divisions such as Treasury) may be harder to assess and is best examined by addressing those questions to the non-executive directors from those divisions.

Finally the Group Risk function was represented at the Group executive level. From approximately 2004 onwards, the Group Risk Director sat on the Group Executive Committee and reported directly to the Group CEO. To the best of my knowledge the Group Risk Director also had full and direct access to the Head of the Audit Committee. Indeed I would stress that the Head of the Audit Committee always encouraged direct feedback from both divisional and Group Risk personnel.

Question 3(3)

The annually-produced five year Business Plan always contained a detailed section on risk management and stress testing (see Question 5 below). The goal here was to make sure the Board fully understood the major potential risks facing the business. In addition, each division would present a strategic update to the Group Board at least once a year and this would include a summary of the major risks inherent in their divisional strategy.

The degree of detail provided in divisional Risk Control Committees was very extensive. I do not believe the problems faced by HBOS were due to a lack of detail provided to the Board about risk factors. With hindsight, at times it may have been the case that the sheer volume of information supplied by every division right across operational risk, credit risk and regulatory risk may at times have made it harder for the Board to fully understand the potential issues facing the business.

Question 3(4)

The link between risk and growth in market share is complex. I think the Board did understand the risks involved in growing market share in retail credit markets and the Board was supportive both of the fast asset growth in retail mortgages in the early years of HBOS and its slower growth from approximately 2005 onwards.

Within Corporate banking, the availability and transparency of market share data is much less clear. As such, the Board would be much more reliant on the individual Risk Control Committees within those divisions and both the Group Risk functions and the divisional executives to monitor the risk profile of the division.

Questions 3(5) and (6)

Each division was responsible for running detailed risk assessments and stress testing models of their portfolios. These stress tests would be presented to divisional risk committees as part of the regular business cycle.

In addition, the annually-produced five year Business Plan (presented to the Board each November) would have a section on risk assessment and stress testing. Clearly I do not have access to all of the annual Business Plans of HBOS and I am trying to remember back many years, but my memory suggests that the core macroeconomic stress testing models were run against a 1 in 25 year economic downturn as agreed with the FSA. Again from memory, in the latter days of HBOS, these assumptions would typically include:

a significant full year deterioration in GDP (ie a full recessionary environment);

a very significant increase in unemployment (for example up by 50%)—interest rates rising to peak at approximately 7%; and

very significant decreases in both residential and commercial property values.

From memory the stress testing would explore the impact of all of the above on impairment levels, and capital levels.

In addition, we stress tested regularly at Group level for the impact of market risk (interest rate and FX), market risk (equity levels) and pension scheme risks.

Looking back at the problems encountered by HBOS and others in the financial crisis, it now seems apparent that much of the stress testing carried out, while valid, failed to highlight the largest risk which materialised in 2007–2008, namely the closure of term wholesale funding markets and the ensuing impact on Corporate banking credit conditions. With the benefit of hindsight, even more of the time spent on both risk committees and Board meetings should have been channelled into examining what subsequently proved to be our biggest strategic weakness, namely our reliance on wholesale funding (see section 6). In hindsight the true level of risk associated with the potential closure of global wholesale markets may not have been fully appreciated.

Section 4. Board Qualifications

Questions 4(1) and (2)

From an executive angle, the HBOS Board always had very extensive banking experience. Indeed the divisions which encountered significant profitability issues were managed by very experienced bankers.

Throughout the existence of HBOS the Corporate division was run successively by individuals with more than thirty years’ experience in corporate banking (in George Mitchell and Peter Cummings respectively). The Treasury division was run successively by Gordon McQueen and Lindsay Mackay (both lifetime bankers). Colin Matthew (who in the latter years ran the Treasury and International divisions) was also a banker of more than thirty years’ experience.

Mike Ellis (who was Finance Director for most of the lifetime of HBOS) was a vastly experienced retail banker.

From a non-executive angle, I do not have ready access to a full comparison of HBOS non-executive banking expertise versus other banks. However Tony Hobson who held the most important non-executive role (that of Audit Committee Chairman) had very considerable relevant experience in the financial services industry, including having been Finance Director of Legal & General for approximately ten years.

In the latter years the Board also included John Mack who had enormous Treasury experience from his career at Bank of America

Clearly a Board by definition can never have too much experience. It seems unlikely, however, given the huge experience of most of the executive team (especially in Corporate, International and Treasury) that the difficulties encountered by HBOS were due to a lack of qualifications.

Question 4(3)

Please refer back to my answers to questions 3(1) and (2) for a full description as to how the central challenge and discipline of divisions operated. The two main sources of challenge to the divisions were from Group central functions (most obviously members of Group Risk/Group Finance/Group Internal Audit), from non-executives present on the Divisional Risk Control Committees and from the Executive Committee.

As an executive, first of all as CEO of the Retail division, I never felt that there was a lack of challenge from the centre to the divisional strategies. Inevitably, however, the strength of challenge was to some degree dependent on the level of expertise in the Group Finance, Group Risk and Group Internal Audit communities.

In the last two years of HBOS’s existence, when I had become Group CEO we had to adjust the “former devolved model” because areas such as funding and liquidity required even greater central co-ordination. To assist in this process, I asked Mike Ellis to return to HBOS in late 2007 as Group Finance Director (he had left approximately three years earlier) as I felt that the Group needed the benefit of his risk management and balance sheet expertise. Though the Group benefited hugely from his return, the core structural problems (particularly the over-reliance on wholesale funding) proved insoluble.

Section 5. The Divisions

Questions 5(1) and (2)

With the benefit of hindsight, the ultimate weakness of the Corporate division proved to be the degree of concentration risk in commercial real estate and in certain large exposures, and the proportion of lower rated exposures. I believe these factors proved to be more important than other issues (such as equity participation).

Whilst the HBOS Retail balance sheet was far bigger than the Corporate balance sheet, even the relatively resilient performance of the Retail assets through the downturn failed to provide a hedge to the Corporate division’s exposure to commercial property. Corporate’s ability to maintain profitability through the downturn was exacerbated further by the lack of a strong deposit base in HBOS’s Corporate banking operations. Without a large SME business, HBOS’s Corporate business lacked a source of deposits from its Corporate division. With the huge benefit of hindsight, HBOS should have pushed even harder to develop an SME deposit base following the inception of the Group in 2001. Whilst building a scale SME deposit business is extremely difficult and takes many years, a bigger SME deposit base would have helped provide an alternative source of funding for the Group and provided the Corporate division with the ability to make margin on both sides of the balance sheet.

Questions 5(3) and (4)

The primary function of Treasury was always to provide funding and liquidity. Even the creation of a large ABS portfolio was to provide an alternative form of assets to be held for liquidity purposes. I believe (though I have not been provided with the relevant documents) that the building up of the ABS portfolio would have started in the relatively early stages of HBOS’s existence (perhaps 2004 to 2005 but this requires verification). In terms of the process for approval, the Commission could consult individuals with the relevant specialist knowledge including Phil Hodkinson (who as Group Finance Director until 2007 over-saw Treasury) and Lindsay Mackay who was Treasurer.

Question 5(5)

When HBOS was created in 2001 it had a very large natural share of the UK mortgage market as the merger brought together five brands (Halifax, Intelligent Finance and Birmingham Midshires from the Halifax business along with Bank of Scotland and TMB brands from the Bank of Scotland business). Three of these brands (Birmingham Midshires, Bank of Scotland and TMB) operated extensively in the specialist (which I understand to mean buy-to-let and self-certified) mortgage lending space.

Over the course of the next two years it became apparent that HBOS’s natural brand and distribution reach was enabling the Group to achieve a very high natural share of the specialist mortgage market. We therefore decided to close the TMB brand and only operate two brands (Birmingham Midshires and Bank of Scotland) in the specialist mortgage space.

HBOS was always aware that specialist lending carried significantly higher risk than vanilla mortgages and we recognised this disparity in relative pricing in order to allow for greater expectations of delinquencies. The degree of specialist lending was apparent to members of the Retail Risk Control Committee.

Whilst HBOS’s exposure to specialist lending would naturally increase its risk profile versus mainstream mortgages, I am not aware of it having led to excessive levels of arrears through the recent recession, having allowed for risk-adjusted pricing. I do not, however, have access to precise data.

Question 5(6)

The lack of International presence (which led to HBOS being perceived as over-exposed to the UK economy) was perceived as a strategic weakness for HBOS from the inception of the Group in 2001. From memory, the UK provided well over 90% of profits in 2001 and a process of internationalisation (albeit from a very low base) began. Whilst asset growth overseas was far higher than asset growth in the UK, I estimate that International profits were still only around 10% of Group profits in 2008.

The International strategy was built principally around the two core economies where Bank of Scotland had international presence prior to the merger (namely Australia and Ireland). The goal in both economies was to grow both Retail banking and Corporate banking. Whilst the decision to internationalise (and thus start to reduce concentration on the UK economy) was a very sound decision in principle, a major problem emerged in execution. It proved easier to expand quickly in both economies in Corporate banking than Retail banking (largely because of distribution strength and the expertise of the local management). This meant that the Group suffered from the downturn in the commercial property markets particularly in Ireland but also, to a lesser extent, in Australia.

As in UK Corporate markets, the Group also found it difficult to attract significant enough deposits in International markets in order to significantly reduce the Group’s wholesale funding requirements (see next section).

Section 6. Wholesale Funding

From the time of the merger in 2001, HBOS did possess a considerable reliance on wholesale funding and this reliance on wholesale funding proved to be the biggest risk facing the Group from the onset of the financial crisis in 2007. The Bank of Scotland had always been very dependent on wholesale funding even prior to the merger whereas the Halifax had been funded largely through a mixture of retail deposits and basic mortgage securitisations prior to the merger.

Through the three or four years following the merger, the Group grew strongly and the requirement for wholesale funding increased further. However the reliance on wholesale funding was recognised as a risk and actions were taken by the then Treasurer (Gordon McQueen) and Finance Director (Mike Ellis) to lengthen the longevity of our wholesale funding profile.

At this stage I was CEO of the Retail division and likewise we worked hard to attract incremental customer deposits to reduce reliance on mortgage securitisations, covered bonds and other forms of term funding.

The monitoring of funding longevity and liquidity risks was largely carried out by successive Treasurers and the degree of over-sight increased further in 2007 following the cessation of term funding markets globally. I was now Group CEO and we introduced a daily liquidity monitoring process under the direction of the Group Finance Director (Mike Ellis) and the Group Risk Director (Peter Hickman). Weekly deposit raising targets were monitored and precise plans were devised to cope with the re-financing of each tranche of long term funding which were maturing.

Whilst all the above describes the degree of oversight being given to monitoring wholesale funding, the fact remains that it was the bank’s reliance on wholesale funding and the collapse of that market which proved to be its fundamental weakness. The biggest lesson of the financial crisis was that many banks (HBOS included) had an over-reliance on wholesale funding and this remained the case despite our attempts to diversify the sources and lengthen the longevity of our funding.

Summary

I have answered all the questions posed to the best of my ability given the time elapsed since events took place and the lack of papers at my disposal. I apologise in advance for any inaccuracies.

30 October 2012