The following are the report's conclusions:

A. Loss determination

Nothing has come to our attention during the course of our investigation thus far that is inconsistent with Allfirst's and AIB's calculation of the amount of losses resulting from Mr Rusnak's fraud. The $691.2 million pre-tax loss figure was the product of AIB and Allfirst's thorough review of Mr (John) Rusnak's actual and fictitious trades, as well as a comprehensive program for confirming the absence of additional unreported open positions. After removing all fictitious trades and figuring the cost of closing out the previously unreported option liabilities, AIB and Allfirst recomputed profit and loss. The loss figure has also been validated by management for reasonableness by comparing the underlying real position over time with actual market positions, and we understand that Allfirst's external auditors, PricewaterhouseCoopers, will give an unqualified opinion on the Allfirst financial statements that incorporate the loss.

B. Responsiveness and co-operation of AIB and Allfirst

The AIB Group, including Allfirst personnel, deserves credit for having taken immediate corrective and responsive action after discovering the fraud. In particular:

•The Group moved immediately to estimate the losses and to inform the market about the losses.

•The Group promptly suspended individuals directly involved in the relevant operations of Allfirst's treasury pending a full review. At the same time, the Group assumed day-to-day management of the Allfirst treasury.

•The Group recapitalized its US affiliate to meet the capital ratios associated with "well capitalised" status.

•The Group has also been vigilant in its efforts to ensure the continued strength from a liquidity perspective of both Allfirst and AIB.

•The Group hired independent counsel and investigators to obtain an independent and thorough review of the problem and make recommendations.

•The Group has begun a careful review of all control and trading activities to ensure the ongoing safety of the institution. AIB Group and Allfirst personnel, including the chairmen and chief executive officers of both institutions, have fully co-operated with our review, and made it explicitly clear that we were to investigate fully and to follow the investigative trail wherever it might lead. And the Group put at our disposal a talented team from the home office of traders, finance personnel, auditors, and risk management personnel, all of whom worked exceptionally long hours at our direction. During our review we found no reason to question AIB's and Allfirst's underlying financial strength.

C. Why the loss was not uncovered and why it was able to grow

There are a number of reasons why the fraud occurred and why it was not discovered for a period of years. We list the principal ones below:

1The architecture of Allfirst's trading activity was flawed. The small size of the operation, and the style of trading, produced potential risk that far exceeded the potential reward:

•A small proprietary trading operation is unable to be sufficiently profitable to justify the sizeable cost of the high-quality risk management infrastructure needed to control risk.

•The style and scope of Mr Rusnak's proprietary trading was quite removed from the customer-generated activity. His trading became essentially a hedge-fund operation. But one lone trader in Baltimore has no competitive advantage in this highly competitive, sophisticated market where many of the other players have specialised knowledge, scale, diversification and specialised expertise.

•Mr Rusnak traded on his own, not as part of a team. Trading activity performed in a team environment is more likely to be controlled. Group ethos and attention lessen the chances that one single trader will engage in unauthorised or unethical activity.

•Aggressive compensation structures are not the ideal way to attract and compensate traders. The structure of Mr Rusnak's compensation may have had the effect of encouraging greater risk taking on his behalf.

2Senior management in Baltimore and Dublin did not focus sufficient attention on the Allfirst proprietary trading operation. There were four principal reasons why the operation did not receive any significant attention by Allfirst or AIB:

•The area was small in terms of expected profits and formal risk limits. The VaR limit (Value at Risk - a statistical measure estimating the maximum loss on a portfolio) for all foreign exchange trading was never more than $2.5 million. Mr (Bob) Ray ( head of treasury funds management and Mr Rusnak's direct boss) gave Mr Rusnak a monthly stop loss of $200,000. The budgeted annual revenue for all foreign exchange trading ran to about $1 to $2 million.

•The Baltimore proprietary trading activities were not part of Allfirst's core business; Allfirst is a retail bank with a strong branch network, sizeable consumer and commercial loan book, and a significant trust business.

•The proprietary trading was under the supervision of a well-regarded former home-office AIB senior manager who had extensive experience managing a much larger trading operation for AIB.

•The data that would have alerted senior management to the size of the problem was altered by Mr Rusnak so that it was masked in the aggregate, periodic reports that senior management typically received.

3 Mr Rusnak was unusually clever and devious. He was thus able to take advantage of the lack of attention to his activities in a number of ways:

•Using his extraordinarily firm grasp of, and influence over, the bank's systems and procedures, Mr Rusnak was able to devise devious ways to obscure his position and profit-and-loss.

•Mr Rusnak took advantage of weak and inexperienced employees in the treasury control groups. These employees, by virtue of their inexperience, poor training, poor supervision and, in some cases, laziness, facilitated Mr Rusnak in circumventing controls.

•Mr (Dave) Cronin (Allfirst treasurer) believed that Mr Rusnak was fundamentally a person of good character, which gave him more latitude in accomplishing his fraud without greater scrutiny by Mr Cronin. Mr Rusnak regularly attended church with, and participated on the parish school board, with Mr Cronin. They both lived in the same town.

•When required, Mr Rusnak was able to use a strong personality to bully those who questioned him, particularly in Operations. He also was advantaged by Mr Ray's strong personality in defending him. Mr Cronin placed an enormous amount of trust in Mr Ray, which Mr Rusnak well understood. Mr Cronin and Mr Ray tolerated numerous instances of serious friction between Mr Rusnak and the back-office staff.

4 Mr Rusnak's activities may also have been facilitated by individuals at other firms. Though we have not had sufficient time thus far to fully delve into this area, we do think the following points are clear:

•Mr Rusnak was able to cloak the scope of his trading activities from Allfirst by trading largely through two significant prime brokerage accounts. Underlying spot trades through the prime brokerage accounts were not recorded by the back office.

•There were certain types of trades made though the prime brokerage accounts that appear unusual and not in accordance with regular market practice. These trades may have aided Mr Rusnak in concealing his losses and his fraudulent conduct.

•Mr Rusnak also was able to fund his trading operations to a significant extent, especially in 2001 by writing deep-in-the-money options with his active counterparties (including the prime brokers). As already noted, these transactions were essentially synthetic loans disguised as derivatives.

•The size and scope of Mr Rusnak's trading activities would certainly have appeared unusual to anyone paying attention at the counter-party firms. Notably, Mr Rusnak would likely have been one of their biggest prime brokerage customers. To date,we have only a limited record of the prime brokers communicating or attempting to communicate with personnel at Allfirst at a level above Mr Rusnak.

5 Treasury management weaknesses at Allfirst also contributed to the environment that allowed Mr Rusnak's fraud to occur. In particular:

•Despite his considerable experience in foreign exchange trading, the Allfirst treasurer appears to have focused virtually his entire attention on interest rates and the overall positioning of the bank's balance sheet, and did not pay the degree of attention that he should have known was necessary to the foreign exchange trading area.

•Having a senior manager who was placed in the position of Allfirst treasurer by AIB, and who maintained close ties to the AIB Group created a degree of unintended ambiguity as to who was really monitoring the adequacy of the job he was doing.

•This de facto dual-reporting structure obscured accountability of the business line. Dublin thought Baltimore was looking after the Allfirst treasurer, Mr Cronin, and vice versa. The Allfirst treasurer turned out to be a key weak link in the control process.

•The Allfirst treasurer directly reported to a number of different Allfirst senior management officials during his 10-year-plus tenure in his job; many of those officials had little or no direct experience overseeing complex treasury operations.

•The Allfirst treasurer was the chairman of the Allfirst ALCO ( Asset and Liability Committee), the divisional representative on the AIB Group ALCO, and a member of the AIB Group's Market Strategy Committee, but neither committee appears to have engaged in any rigorous examination of his conduct in supervising the foreign exchange trading at Allfirst. As noted above, the reports that were presented to the Allfirst ALCO used aggregated data that masked Mr Rusnak's activity because Mr Rusnak had manipulated the underlying data.

•The manager directly supervising Mr Rusnak, Mr Ray, contributed significantly to the environment that allowed Mr Rusnak's fraud to occur, and indeed, was the ideal foil for Mr Rusnak's scheme, because he:

•never took the time or made the effort to understand the foreign exchange trading business;

•discouraged outside control groups from gaining access to information in his area; and

•reflexively supported Mr Rusnak whenever questions about his trading arose.

6The proprietary currency trading business was inadequately supervised. This conclusion inevitably follows from a variety of facts:

•Other than Mr Cronin, there was no one in Baltimore who had the experience or expertise to deal with a proprietary-currency trading operation.

•Mr Rusnak's chief supervisor for much of the period, Mr Ray, was a former bond trader who took little interest in the currency trading area and, once that area was assigned to him, he did not obtain the necessary expertise to supervise the area.

•The reporting of front, middle and back offices to Mr Cronin was inconsistent with maintaining an appropriate separation of duties. The proper risk management structure would have had operations and, perhaps the middle-office, report independently through a chain to the CEO through, for example, the CFO or risk assessment chains.

•This increased reliance on the treasurer, someone whose focus on the risk controls might have been clouded by his responsibilities to the front office. Internal audit and risk assessment never commented on this flaw. More specifically:

•While the back and middle offices officially reported to the Allfirst treasurer, Mr Cronin, they seemed to be influenced more by the head of treasury funds management, Mr Ray. Indeed, at one time, these individuals reported directly to the head of treasury funds management. Later, in response to a supervisory recommendation, their reporting line was switched to the Allfirst treasurer. Whenever a conflict between the front and back office arose, Mr Cronin too often deferred to Mr Ray.

•Additionally, in response to general efforts to reduce expenses and increase revenues, the Allfirst treasurer permitted the weakening or elimination of key controls for which he was responsible. As noted above, senior management should not have had these control groups reporting to the Allfirst treasurer, who was also responsible for producing profits and losses. Mr Rusnak was able to manipulate this concern for additional cost cutting into his fraud.

•The head of treasury funds management inexplicably ignored numerous warning signs of Mr Rusnak's activity. He did not closely review daily profit-and-loss statements. He did not question excessive daily volumes. He largely ignored warnings from Operations. To make matters worse, he also made it very difficult for risk oversight management personnel, and non-trading staff, within Allfirst to obtain information necessary to do their work.

•Mr Cronin was a "hands off" manager. He focused his attention on interest rates and the overall positioning of the bank's balance sheet and largely ignored foreign exchange trading. He placed too much faith in the strong-willed head of treasury funds management who had no prior foreign exchange experience. He did not display an adequate appreciation of signals that there were serious issues behind front and back office friction.

•Neither Mr Cronin nor Mr Ray did enough to understand and support the principles of strong operational procedures and controls in managing their business. They demonstrated a lack of interest in these details.

•In addition to their failure to understand the details of Mr Rusnak's activities, both Mr Cronin and Mr Ray missed the big picture. It appears that they did nothing to gain an understanding of active leveraged risk taking. This is evidenced by the off-market compensation arrangements and the lack of trading capital controls.

•Messrs Cronin and Ray failed to perform overall "reasonableness tests" of Mr Rusnak's activities. The size of gross positions, the level of daily turnover and the extent of broker attention were excessive given Mr Rusnak's expected and budgeted P/L ( profit/loss) and his VaR limits.

As an example, even if Mr Rusnak's purported hedge strategy existed, the basis risk on the visible gross size of the positions would have produced P/L swings that were many times greater than Mr Rusnak's stop loss limits. Mr Ray should not have missed this obvious point.

•Furthermore, Mr Cronin and the head of treasury funds management never requested sufficiently detailed risk assessment reports, which would have facilitated a greater understanding of the risks being managed.

•Mr Cronin did not adequately "manage up". For example, he ordered that Mr Rusnak's trading be temporarily discontinued on two occasions (a serious action). He did not mention these suspensions to his management.

•Mr Cronin allocated too large a portion of his resources (his time, expertise of his operations people, etc.) to the interest-rate portion of his business. Mr Cronin perceived his responsibilities to be mostly interest-rate oriented.

7 AIB Group Risk, AIB and Allfirst senior management groups and the respective Boards assumed that the control and audit structures governing the trading activities that were being conducted at Allfirst were sufficiently robust.

•A clear lesson of this situation is that proprietary trading (and perhaps all trading) activities are an extremely high risk activity no matter how small the activity appears to those in the senior levels of a bank's organization.

•With the benefit of hindsight it is clear that:

•The AIB Group Risk function should have had a greater role in supervising, monitoring and auditing the trading activity that was taking place at Allfirst;

•The ALCO and other relevant committees that reviewed the treasury functions at Allfirst and on a Group-wide basis should have received better information concerning the nature of the proprietary trading activities occurring at Allfirst;

•The senior management teams at both AIB and Allfirst should have insisted on much more rigorous risk assessment and audit reviews relating to such activities;

•The audit committees of the Boards of Directors of AIB and Allfirst should have been better informed of those risks and should have received better reporting as to how those risks were being supervised, monitored and audited.

8 Risk reporting practices should have been more robust.

Specifically:

•The risk-reporting practices above the Allfirst treasurer, Mr Cronin, were not uncommon for entities of the size and characteristics of Allfirst and AIB.

•The Allfirst ALCO, while not aware of many warning signs, failed to act on those they did see.

•Additionally, ALCO at Allfirst appears to act passively by addressing only what they were presented, instead of proactively reaching out to those taking and controlling risk. ALCO should have had all major market/interest rate risk taking areas in the organisation present to them from time to time.

•Allfirst risk management groups appeared to place excessive reliance on the VaR model and ignored other risk information that was available to them. ALCO should have requested a presentation on the strategy by the head trader so they could evaluate the activity themselves.

Additionally, ALCO should make use of scenario analysis and stress testing in addition to VaR. These methods highlight "long tail" or "event" risks which, while not probable, can be ruinous.

•AIB appears to have placed far too much reliance on the structures in place at Allfirst and on the reliability of Mr Cronin. Despite the fact that AIB recognised that Mr Cronin was the only employee at Allfirst with extensive experience managing or supervising a proprietary trading operation, and despite the knowledge of the AIB Group Risk that Mr Cronin appeared to be relying heavily on AIB guidance as it related to treasury policies and procedures, no proactive measures were taken at the Group level to actively monitor this area. At a minimum, Mr Cronin's participation on the AIB ALCO provided an opportunity for more active scrutiny of what was going on in Allfirst treasury than was availed of.

9 No policy and procedures review

There was no attempt made to benchmark the policies and procedures followed by the foreign exchange operations department or the ALM(asset and liabiilities management) /Risk Control department against "best practices" or the foreign exchange Committee's guidance, although the Funds Management Policy was reviewed each year by the Allfirst board. Indeed, we saw no evidence that risk management or treasury officials periodically reviewed these policies and procedures for changes in the business or the markets, beyond periodic consultations with Group Risk regarding the policies employed at AIB.

10 Scrutiny of Mr Rusnak's employment arrangement was inadequate.

The structure of Mr Rusnak's compensation arrangement was, from a risk management perspective, questionable.

11 Failure to appreciate regulatory concerns

Prior to 1998, Allfirst was supervised by the Office of the Comptroller of the Currency(OCC). On several occasions, the OCC raised some concerns about risk management in the foreign exchange trading area. The OCC's concerns included:

•Foreign exchange traders exceeding vega limits.

•Inadequate level of information provided to ALCO.

•Lack of a review of foreign exchange trades for off-market prices.

•The fact that operational and management reporting lines and the business reporting lines funneled into Mr Ray. In hindsight, it does not appear that the company gave these OCC concerns the full attention they deserved. It should nonetheless be noted, however, that the context in which these concerns were raised was one in which the OCC had generally indicated that risk management of trading and derivatives activities is strong.

12 A flawed control environment existed. Allfirst's treasury operations department was deficient in a number of respects. The deficiencies in the operations area included:

•Treasury operations personnel lacked experience and expertise. Allfirst's treasury operations department was staffed by personnel who did not have the experience required for the nature of the trading activities undertaken. Examples of this inexperience include:

•No one questioned why option counterparties repeatedly did not exercise profitable option transactions, even though automatic exercise of in-the-money option positions is customary.

•No one questioned why two identical options with different expiry dates would have the same premia.

•No one questioned with senior management the large gross foreign exchange activity, even if such activity was relatively small on a net basis.

•No one reviewed the large daily P/L swings reflected on the general ledger, which would consistently net down by month end.

•The practice of not confirming certain types of trades was not questioned for a very long period of time.

•Members of the treasury operations staff were inadequately trained and supervised.

•Certain treasury operations personnel exhibited careless behavior. There existed a combination of inadequate written procedures, a failure to follow those procedures that did exist, and a propensity to modify practices at will.

•Traders had excessive influence over operating procedures.

•The confirmation of trades, a fundamental process in the business, was haphazard and very badly managed and executed. When the failures in the confirmation process in the Prime Accounts became evident, there was no review undertaken of all confirmations.

•The independence of the pricing process for revaluing open foreign exchange contracts was compromised.

•Daily management reports were not independently verified.

•Remote access to bank systems was not controlled. The monitoring that existed was never checked.

•There was a failure to reconcile and settle broker bills (this function was actually conducted by Mr Rusnak himself).

•There was no validation of counterparty details and there was undue influence of traders on the counterparty set up process.

•There was no independent Treasury finance function.

•There was a lack of segregation of duties.

•There were instances of password sharing.

13 ALM/risk control deficiencies

There were also deficiencies in the ALM/Risk Control area as well at Allfirst:

•Daily stop-loss P/L was not reconciled to the general ledger.

•ALM/Risk Control failed to understand the risk of the products traded and the strategy undertaken. They failed to produce reports to support the management of these risks.

•ALM/Risk Control constructed a reporting framework which was not based on independent information but relied instead on inputs from Mr Rusnak.

•There was no reconciliation between the month-end P&L figures and the GFA-sourced ( Guidelines for Foreign Exchange ) monthly foreign exchange income report.

•Inadequate supervision and review of the work of the risk analyst who produces the daily VaR and stop-loss information.

14 Credit risk deficiencies

There were deficiencies in the credit risk management area:

•Oversight of the counterparty credit risk management process was inadequate.

•There was undue reliance on the trader to establish the status of a counterparty. The impact of this is that Allfirst treasury repeatedly entered into transactions with counterparties for whom there were no valid limits.

•Repeated breaches of limits were not properly followed up.

15 The risk assessment group failed to adequately review the control environment in treasury. In particular:

•Internal audit and risk assessment failed to adequately scrutinise the control environment in Allfirst Treasury and failed to exercise sufficient diligence in examining Mr Rusnak's trades and audit trail.

•Allfirst Treasury and those responsible for auditing its activities and implementing appropriate risk management practices failed to apply readily available guidelines for best practices, including the Guidelines for Foreign Exchange Trading Activities published by the Foreign Exchange Committee of the Federal Reserve Bank of New York.

•While the risk assessment group commenced a Treasury Credit review in 2000, this exercise has yet to be completed.

•To the extent the risk assessment group had substantive concerns it did not take sufficient responsibility for ensuring that such concerns were adequately addressed.

•The quarterly Risk Assessment Reports did not adequately document the weaknesses described in their own working papers andpersonal records.

•The failures and warning signs that were raised (e.g. the issue of possible trader influence on market prices used in valuing positions) should have been more aggressively pursued and resolved more quickly.

•The quality of the work done by internal audit in this area did not demonstrate a clear understanding of the risks associated with the business strategies. Key controls were either not tested, or the testing undertaken was not effective. The review process did not identify the shortcomings in the two audits in the foreign exchange operations area as it related to the prime brokerage accounts and processing on the DEVON system.

•The culture of Allfirst's risk, operations and internal audit functions, as applied on the whole, to the Allfirst treasury operations, was too deferential to the business lines. It is true that individuals in these functions were bullied from time to time by Mr Ray and Mr Rusnak.

16 Mr Rusnak was allowed to trade on vacations.

It is for good reason that banks in the United States are required to have a guideline that bars traders from trading two weeks per year. Having a second employee take over for even the most trusted bank employee is a mechanism for uncovering fraud.

Although Mr Rusnak did take vacations, the efficacy of the two-week rule was vitiated because Mr Rusnak was allowed to continue trading from his computer. He also traded at home and at night, but this trading was not monitored appropriately. The control weakness was the failure to verify this after-hours trading.

D. Conclusions

As our conclusions make rather clear, there are substantial reasons to be highly critical of many of the individuals within Allfirst treasury. At the same time, however, we have no definitive basis to conclude at this point that anyone within Allfirst treasury other than Mr Rusnak had actual knowledge that Mr Rusnak was engaged in fraudulent or improper trading activity, before the events leading up to the discovery of the fraud in February 2002. For that matter, as noted above, nothing has come to our attention during the course of our review that indicates that anyone at AIB or Allfirst, outside of the Allfirst treasury group, were involved in, or had any knowledge that, fraudulent or improper trading activity was occurring at Allfirst, before the discovery of the fraud. There has been no effort of which we are aware by senior AIB or Allfirst personnel to conceal or cover up any facts.

Additional Work

The short time frame of our review has not permitted the team to produce answers to a number of important questions regarding the loss. Accordingly, we recommend that the AIB and Allfirst boards perform certain additional investigative, legal and ancillary work along the following lines:

1. The boards should determine the potential role of prime brokers or other trading partners.

•As mentioned above, there are instances uncovered during the investigation where we found transactions with prime brokers that were off-market or not in accordance with "market standards". It appears that these digressions may have helped Mr Rusnak conceal his positions and/or his losses.

•"Unusual" trades should be fully reviewed to ensure that no impropriety took place and to clarify fully whether the prime brokers or other dealers, or personnel at these firms, obtained any benefits in their dealings with Mr Rusnak other than full bid-offer spreads and market fees.