Adopting a deferred pay scheme for bank managers would provide them with needed funding during a downturn and would incentivize more conservativism when it comes to risk-taking.

In 1929, Goldman Sachs was almost brought down by Waddill Catchings, a manager-partner.

Catchings had created the Goldman Sachs Trading Corporation, essentially a trust that used debt to buy companies that had themselves used debt to buy other companies. When the crash came, the Goldman Sachs partners agreed to place about half of the partnership’s capital of $20 million in the venture to guard against the company’s failure. Thus, the adverse consequences of Goldman Sachs Trading Corporation’s activities were largely borne by the manager-partners. Goldman’s problems, if not solved by the firm’s internal capital, could have jeopardized the welfare of the broader financial system. The lesson might be relevant to the current banking crisis.

Early in March of this year, Silicon Valley Bank (SVB) tried to raise $2.25 billion by issuing shares after reporting a loss, but its effort failed. The aim was to plug a $1.8 billion hole in its balance sheet caused by the sale of $21 billion worth of bonds at a loss. The bank’s share price had lost nearly 60% of its value, according to the press.

The inability of SVB to raise capital is not unique in US (investment) banking history. Early in 2008, Lehman Brothers had a similar experience. During the savings & loan crisis, the Bank of New England, a large regional bank, could not raise capital for its shares almost at any price. Failure of seemingly healthy firms managed by competent executives inflicted a significant cost on stakeholders in the US economy and beyond.

The experience of these firms emphasizes the importance of risk management in the banking industry. It is a fast-changing world, and eventualities are rarely anticipated not only by bank management and investors in financial firms but also by bank regulators and supervisors. The SVB case shows that current rules and guidelines are just baselines for addressing standard problems and not for unanticipated events that seem to be surfacing at an increasing frequency. 

Given a low likelihood of raising funds from investors, individuals, and institutions when banks are in financial difficulty, the question that arises is: Could banks raise funds internally? That is, could they use their employees as a source of capital, similar to how the partners in Goldman Sachs provided needed funds in 1929? Because traditional banking firms do not have partners, who then can or should provide capital within the firm when funds are needed quickly? The scale of some banking firms is large, thus insiders could not possibly cover the capital shortfall quickly enough in the event of a shock to mitigate the loss of value and restore stakeholders’ confidence.

Few in any bank have enough liquidity to supply funds during a liquidity crisis. Could SVB have created an account to act as a liquidity buffer? If so, how big should that buffer have been? SVB (and other banks) could have created an account funded by deferred pay by the bank senior management and managing directors. A reasonable fraction of each senior manager’s annual compensation could be deferred for a predetermined period and would be subject to vesting requirements after the end of the deferral period. The scheme would generate a pool of funds to be utilized when needed. The size of the deferral buffer, given the bank’s total compensation cost, depends on the rank and number of employees required to defer their pay, the fraction of annual pay that is deferred, the length of the deferral, and the vesting period. The idea presented here follows Acharya, Mehran, and Sundaram (2016) and Mehran and Tracy (2016). A similar idea was advocated by the Squam Lake Report.

The table below shows SVB’s total annual compensation and benefits over the eight years 2015–2022, total bank employees measured by averaging over four quarters employment numbers for each year, and average pay for bank employees. Mehran (2022) states that 90% of total pay in the three largest US banks was distributed to 20% of their employees. To be conservative, I assumed that 60% of total pay in SVB is awarded to managing directors and senior management, column 5. I also assumed that 40% of pay to senior management is deferred, column 6. The time frames are a deferral period of four years and uniform vesting also over four years.

1234 (2/3)5(0.6×2)6(0.4×5)
YearTotal compensation (millions $) Employees         Average pay ($)  Management
(millions $)
Deferral (millions $)
20154722,000                 236,000             330                                           132
2016509    (7%)         2,237 (12%)       228,000             356                                           143
2017604    (17%)      2,399 (7%)         252,000             423169
2018725    (20%)       2,718 (13%)       267,000             508                                          203
2019986    (36%)       3,213 (18%)       307,000             690                                           276
20201,215 (23%)  4,123 (28%)       295,000             851                                           340
20212,120 (74%)       5,782 (40%)       368,000             1,490                                       596
20222,296 (8%)         8,010 (38%)       287,000             1,607                                       643  
Source: Regulatory filings.

The table shows a large growth in total pay and employment starting in 2019, with a smaller growth in pay in 2022 (the compensation numbers are not inflation-adjusted). The latter could be explained by the recruitment of low skilled employees in 2022 and by bonus deferrals from 2022 to 2023. SVB average employee pay is significantly larger relative to average employee pay of the three largest US banks (see Mehran, 2022). For example, average employee pay in 2021 was $368,000 for SVB and $143,000 averaged over Bank of America, Citigroup, and JPMorgan Chase. Large banks have thousands of branches and hire a significant low skilled labor force who are paid relatively less. Further, branch pay is likely correlate with pay in the local market, which is generally below the California pay scale. That being said, other explanations may exist for the differential between the average pay at SVB and the three largest US banks.

Calculation of total deferrals at the end of each year is reported below:

2015–2018: 132 + 143 + 169 + 203 = 647 ($ millions)

2019:  647 + 276 – 0.25 x 132 = 890 ($ millions)

2020:  890 + 340 – 0.25 x 132 – 0.25 x 143 = 1.161 ($ billion)

2021:  1,161 + 596 – 0.25 x 132 – 0.25 x 143 – 0.25 x 169 = 1.646 ($ billion)

2022:  1,646 + 643 – 0.25 x 132 – 0.25 x 143 – 0.25 x169 – 0.25 x 203 = 2.178 ($ billion)

The $2.178 billion deferral buffer is very close to the amount that SVB attempted to raise. Thus, funded deferred cash bonuses would function in much the same way as partnership capital if the risk takers are required to set aside a fraction of their pay every year to manage the bank’s need for capital during a crisis. Banking firms or regulators could decide on various features of the scheme. Regulators most likely would opt for a uniform policy across asset size for the large banks. Bankers most likely prefer the current pay policies.

A few points should be noted:

Firstly, the deferral scheme offers a way for a bank to have access to funds needed in a time of crisis and avoid going to the market as disclosure has its own consequences, as evident in the case of SVB. This does not necessarily suggest that SVB could have survived if it had a buffer of $2.178 billion.

Secondly, if it already had existed, the scheme most likely would have generated a bank different from the failed bank. Creation of the deferral buffer could induce conservatism. Thus, focusing on the size of the buffer misses the potential benefit of a pay deferral scheme. Bankers seeking to protect their deferred claim are more likely to make judicious investment, capital structure, and payout decisions.

Thirdly, if the 30 largest banks had adopted the scheme in 2015 or earlier, the system could have had access to  billion to $100 billion in liquidity at the end of 2022.  This is about 1/6 of the funds available under the Troubled Asset Relief Program capital infusion during financial crisis.

Fourthly, while it may seem that no amount of internal capital contribution could be sufficient for banks facing scenarios such as during the 2007–2009 financial crisis, funded deferred pay scheme could have provided a life-line to banking firms during crisis. For example, Lehman Brothers could have access to just over $11 billion cash at the end of 2007 before its crash in 2008. The bank grew its works force by 10% to 25,866 in 2007. Its average employee pay in 2007 was $332,226, average pay for the largest three banks in that year was $165,000.

Next, just as during the financial crisis, a number of commentators have argued for raising more equity capital since the failure of SVB. While their concerns have merit, more capital does not produce conservatism on the part of management and does not improve bank governance. Thus, pay deferral could complement the benefit of a higher level of equity capital to enhance financial stability.

Additionally, bankers might consider the scheme voluntarily. The US economy is evolving, and drivers of some risks are unknown. Regulators might inflict risk on the system. Thus, bankers might want to protect their own employment contract in the industry by producing a safer bank. Pay deferral could contribute to that aim and, at the same time, the decision would be a step toward shared interest with the public.

A number of people has advocated for new regulations and supervisory improvements. Reforms often prepare the public for crises or problems, but they are rarely incentive-compatible (see Mehran, 2021, for an example). The pay deferral scheme is in effect an incentive-compatible safety tool.

The pay deferral scheme is not likely to have adverse labor market consequences (see Mehran and Tracy, 2016). This does not imply that labor would not take a risk if a funded deferral pay scheme is adopted, this is so if pay is tied to performance.

A typical narrative about past banking crises says that the problem was the failure of bank management. Terms such as “misconduct” and “abusive” are used along with other terms such as “excessive risk taking” to conclude that a contributing factor to a financial crisis was management compensation. To support this narrative, Mehran (2022) indicates that a Google search of “bankers’ pay and financial crisis” produces a few hundred million hits. Similar sentiments formed immediately following revelation of SVB’s failure.  

For a few decades, academics have made a case that banks or banking firms are different. While regulators and supervisors can play an important role in bank governance, the evidence points to their ineffectiveness regarding overseeing bank leadership. Thus, the governance and discipline of banks are left to bank stakeholders who have little power to bring about a change when needed.

Further, regulators contribute to the weakening of market discipline by reassuring the public that they are in control of the banking sector. The conduct of regulators has the potential to reduce the incentive for information production that could discipline the banking firms. For example, in the case of SVB, some of the most sophisticated deposit holders did not notice that the bank had negative tangible equity.

The conclusion is that now may be the right time for banks and their boards as well as regulators to consider an alternative to achieving financial stability. The funded pay deferral scheme could protect the public, shield bankers, and restore regulators’ reputation. Finally, it should be noted that the funded pay deferral scheme idea outlined here was developed by the staff of the Federal Reserve Bank of New York and was shared with the Board of Governors in 2013, and then in 2015 to support the Board’s effort to craft compensation guidelines for the banking industry sanctioned in the law by Dodd-Frank Reform Act. The Board did not consider the idea and the US has not reformed bankers’ pay yet.

Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.