Abstract: The collapse of Silicon Valley Bank was not a failure of its main business but a failure of its asset and liability management. This article summarizes five lessons for China's small and medium-sized banks on asset and liability management.
The sudden collapse of Silicon Valley Bank, or SVB, on March 10, 2023, shocked the world.
This is firstly due to the bank's successes in the niche market of credit financing services for technology start-ups, where it was still the market leader prior to its bankruptcy.
Second, because it does business with more than 50% of American technology firms and receives the majority of its deposits from them, a bankruptcy would have a significant negative effect on technological innovation.
Third, SVB has reached a considerable size after 40 years of development. At the end of 2022, it had total assets of US$211.793 billion, placing it 20th among US banks. This scale is approximately 1.5 trillion yuan, which corresponds to a position outside the top 20 among China’s commercial banks and is neither too big nor too small.
Fourth, SVB has, this time around, sparked a significant amount of deposit runs. The market is concerned that its size could start a chain reaction and lead to systemic risks.
Even more intriguing is the explanation for the sudden demise of a bank with such a stellar reputation in a particular industry. According to what we currently know, SVB 's typical line of business poses no risk from the standpoint of commercial bank operations. But because of its successful business model, it has developed a unique asset-liability structure. The main causes of its collapse were poor asset-liability management and the operational management abilities of commercial banks.
Ⅰ. REASONS AND EFFECTS OF THE BANKRUPTCY OF SVB
First, look at SVB 's liabilities.
Its total liabilities at the end of 2022 were US$195.498 billion, with general deposits making up 88.5% of that amount ($173.1 billion). The deposits peaked in 2021 at $189.2 billion, or 89.2% of the $212.001 billion total assets that year. Its deposits significantly decreased as of 2022. This liability structure is excellent in most situations. Only the rural commercial bank system in China's commercial banking industry has recently had a general deposit-to-total liabilities ratio of about 80%.
According to SVB's annual report, Demand deposits and other transaction accounts in general deposits total US$132.8 billion, while savings and time deposits total US$6.693 billion; 76.72% of the total deposits were made up of demand deposits and other transaction account deposits. This liability structure is terrible. Generally speaking, personal savings deposits, time deposits, and corporate demand deposits are the liabilities of banks that are the most stable. Companies have varying policies regarding the use of funds, and banks occasionally need to be aware of these policies as part of routine management. Allocating assets is extremely difficult when liabilities are unstable.
In reality, banks primarily concentrate on both liability management and asset management to address the volatility of liabilities. On the liability side, it is typically used in interbank lending transactions to hedge momentary fluctuations in deposits. Active liability methods, such as large-denomination certificates of deposit or bonds, can be issued under specific conditions to enhance the maturity structure of liabilities. On the asset side, distribute cash, central bank reserves, high liquidity trading assets, interbank lending, and working capital loans appropriately.
Strangely, SVB usually did not seem to have effectively managed such an unstable liability structure. For example, they did not operate interbank transactions. When there was a run on deposits, they did not resort to interbank lending, issuance of large certificates of deposit and bonds, etc., but directly sold assets and tried to raise funds by issuing additional stocks to make up for the liquidity difficulties caused by the loss of deposits.
Second, look at the assets of SVB.
Among SVB’s assets, loans amounted to $73.614 billion, including $58.459 billion in commercial and industrial loans and US$13.128 billion in real estate loans. Loans accounted for 34.76% of total assets. In addition, it holds $722 million in federal funds, $17.223 billion in Treasury securities, $578 million in federal agency securities, $91.461 billion in residential mortgage-backed securities, and $735 million in other securities. Securities accounted for 52.27% of the total assets. The cash and due deposits are only $9.116 billion, accounting for 4.3% of total assets. If we regard commercial and industrial loans as the core business of SVB, that is, direct loans to support scientific and technological enterprises and bridge loans to venture capital funds, the real main business of SVB's total assets only accounts for 27.6%. But in terms of quantity, its main business is securities investment.
Such an asset structure is completely out of the question for regular commercial banks. Such an asset structure is completely absurd, even when looking at SVB's operating characteristics.
Loans and bonds are both debts, but they are not the same. The lender and borrower are generally clear in loans; most interest rates are determined by adding points to the benchmark interest rate, and the income is essentially determined, less affected by market interest rate fluctuations; loan liquidity is weak; more attention is paid to the borrower's credit risk. Bonds are fully negotiable, allowing holders to make changes. Bondholders' willingness to hold bonds will change in response to their own financial situation or market conditions, which will also affect bond prices. Holders, for example, may sell a large number of assets due to financial deterioration, lowering the price of related assets. Bond prices are affected by changes in market interest rates and can fluctuate at any time due to changes in the issuer's credit risk. Bonds, as a result, concentrate on the issuer's credit risk, market risk, and liquidity risk. The risk logic and management logic of the two are distinct, as are their roles in commercial bank asset-liability management.
Commercial banks generally manage held-to-maturity bonds as quasi-credit business, which follows the process of credit risk management, and most of the time banks do not invest in the bonds issued by noncredit clients. A proper share of available-for-sale bonds can be held as operating assets for profit-making and are often used as an instrument to manage liquidity. It is normal that this portion of bonds may sometimes generate no return or even losses.
In terms of asset management, commercial banks must first cover the withdrawal of deposits. Usually, the excess reserves should account for around 20% of their deposits, including cash on hand and reserves deposited in the central bank. Given the features of SVB’s liabilities, the excess reserve ratio should be higher. But in fact, the cash and time deposits held by SVB only account for 5.27% of the total deposits, which cannot meet the daily demand of deposit withdrawal. The management of cash and time deposits shows that SVB valued profitability and basically did not consider managing the liquidity of its bond position.
Holding such a large share of securities exposed to interest rate risk, SVB seemed to fail to take proper interest rate management, such as interest rate swaps, in the face of imminent rate hikes by the Fed after quantitative easing. During the low-interest-rate period between 2020 and 2021, SVB rapidly increased its securities position. But when the Fed started aggressive rate hikes later, it became too late for SVB to adjust its asset allocation in the second half of 2022.
In 2019, SVB only held 61.8 billion US dollars of deposits; in 2020, the figure reached 102 billion US dollars, up by 65%; in 2021, the figure climbed to 189.2 billion US dollars, up by 85%; in 2022, it declined to 173.1 billion US dollars, down by 8.5%. In just three years, deposits rose by 180%. In the same period, loans increased from 32.9 billion to 73.6 billion US dollars, up by 123.7%. But the growth of loans is far lower than deposits. The surge of deposits occurred at the same time as the Fed’s quantitative easing. But since Ben Bernanke, every chair of the Fed had been looking for chances to increase rates. In the second half of 2021, the Fed started to wind down bond-purchase. Subsequently, the deposits of SVB dropped, but SVB did not take any measures to manage its assets or liabilities.
Let’s take a closer look at the customer base of SVB.
SVB has nearly 37,500 deposit customers with an average balance of 4.616 million US dollars. The customer composition is highly concentrated, mainly including tech start-ups or venture capital firms. When providing services to these customers, SVB only focused on corporate business and did not provide retail banking business to the employees of these companies. In fact, the employees, especially the management, all have a high income. If SVB could provide services for these people, it could have at least improved its liability structure.
In terms of capital flows, these companies have their own features and operate in a very different way from normal businesses.
When a venture capital firm raises funds, a large amount of money flows into the account and becomes corporate deposits. When the firm finds an investment project, it will spend the money and thus reduce corporate deposits. There is no pattern in the timing or the scale of investment. When the firm successfully exits the project, a massive amount of money flows in again. In most cases, these firms do not have operating cash inflows. Their deposit accounts tend to rise and fall sharply with great volatility and no pattern.
As for tech start-ups, a large amount of money is credited to their accounts when they get investment. With their development, they will withdraw the money. The withdrawals have a certain pattern, and the operating cash inflow is rather low. Thus their deposits will decrease gradually until the next round of financing is secured.
As can be seen from the features above, the source of SVB deposits comes from private investment in tech start-ups. When the Fed started to aggressively raise interest rates and tighten policy, the source of investment in tech start-ups contracts. As tech start-ups and venture capital firms only saw capital outflows without inflows, deposits in SVB shrank rapidly.
Based on its annual report, the management of SVB did notice the features of its deposits. But the rapid expansion of deposits over the past three years was not a result of the bank’s active operation but a result of passive increase. Because of this, the growth in deposits did not lead to growth in loans in the same proportion. Instead, the added deposits were invested in high-quality securities with high liquidity. If the deposits were used to grant loans, it would cause real maturity mismatches. But investing in securities is an act of asset allocation rather than an act of balance sheet management, which neglects market risks and liquidity risks.
In summary, there is no problem with the SVB’s asset quality or asset liquidity. In other words, the reason for the collapse of SVB is not its asset quality, or the mismatch between assets and liabilities, but its mismanagement of its overall balance sheet. When developing a unique liability structure in its unique operation, it can manage credit risk in its unique deposit businesses but is unable to manage the balance sheet risks outside of its unique businesses.
Although the collapse of SVB shocked the world, it will not have great implications for technological development. Silicon Valley Bank is only a small investment bank uniquely dedicated to serving innovation-driven tech companies, with only 60 billion US dollars flowing into the tech industry. Therefore, it is not the main financial driver of technology investment. Tech start-ups also have a feature. The success of a tech start-up will have a huge influence on the society and economy, but one failure will barely have any impact. In this sense, the shutdown of SVB will affect existing tech start-ups, but will only have limited direct impacts on the whole society and economy. As the assets held by SVB are of high quality, the change in market interest rates will only cause losses instead of asset liquidation. Depositors will have limited, if any, losses. However, due to the public nature of banks, the failure of SVB might undermine the creditworthiness of small and medium-sized banks, which could lead to systemic risks.
II. FIVE TAKEAWAYS FROM THE SILICON VALLEY BANK COLLAPSE
First, to do businesses with special features, commercial banks must choose customer groups and industries that boast sustainable development.
It is a very good development path for small and medium-sized commercial banks to expand particular customer groups and provide special services based on their own endowments so as to form distinctive core competitiveness. However, the selection of customer groups or industries must take into consideration of sustainability. If the chosen customer group and industry have obvious cyclical features, it will sow hidden dangers. Some banks in China once set up industry-specific divisions, but when the chosen industry entered a downward cycle, these divisions became unsustainable. In the case of Silicon Valley Bank, the tech industry is generally sustainable despite some fluctuations.
Second, the way that the Silicon Valley Bank supported tech start-ups is respectable and worth learning.
Current information available shows that Silicon Valley Bank's core business of supporting tech enterprises has no problems at all. Its efforts in this area in 40 years have proved successful and are worthy of learning. The key to the issue is the model design to cover losses resulting from credit risk of high-tech companies. First, Silicon Valley Bank has strict selection conditions for loan recipients; second, loan risks are hedged not only by high-interest rates but also by other charging methods; third, professional cooperation between banks and venture capital institutions is the key. These practices, objectively, require a corresponding legal system and market atmosphere. Subjectively, banks have to equip themselves with professionalism, patience, quality, and capability, rather than pursuing speed and scale blindly.
Therefore, the practices of the Silicon Valley Bank cannot be completely copied by banks in China. Some regions and banks should be encouraged to conduct pilot projects to explore desirable business models, charging models, and accounting models to cover the loss caused by credit risk of tech enterprises.
Third, while forming featured business capabilities, commercial banks should never loosen their assets and liabilities management.
The bankruptcy of Silicon Valley Bank this time was not a failure of its main business, but a failure of asset and liability management. Despite the fact of conducting both deposit and lending business, people often regard banks’ risk management capability as the ability to manage credit risk because of the critical importance of the quality of loans. In reality, a bank's operation relates to all aspects, while credit risk management alone does not guarantee a bank's safe operation. Most bank failures in history were triggered by liquidity risk. This is also a major lesson that Silicon Valley Bank has taught us.
In addition, while conducting featured business, it is important to explore more business opportunities around the core business to offset the concentration risk. Some small banks in China focus on serving small and micro enterprises, and in addition to providing professional credit services, they also provide retail services to corporate employees and nearby customers so that the customer structure, business structure, and asset and liability structure have a better balance in addition to their featured business. On the contrary, some banks' industry-specific divisions tend to focus their business solely on corporate assets, making no efforts to balance liabilities, attract different customers and diversify businesses.
Fourth, it is not recommended to set up a separate specialized bank to serve high-tech enterprises, but to encourage branches of large banks to provide specialized support for them, while asset and liability risks can be managed at the head office level.
The reason for the collapse of Silicon Valley Bank this time can be attributed to the excessive concentration of customer types on the surface. However, the concentration of customer types is inevitable for specialized commercial banks as such concentration is both a feature and a core competency. It is the special customer base that creates a special liability structure, which makes asset and liability management difficult. Objectively speaking, it is extremely challenging for a bank to maintain its liquidity, safety, and efficiency in the face of such a high proportion of highly liquid deposits. Therefore, if China wishes to introduce the model of Silicon Valley Bank, it can borrow the bank’s specific business model, but not the operation model. In other words, not to set up a specialized bank to do the business.
Fifth, when encouraging banks to support the real economy, it is always important to put the safe operation of banks' assets and liabilities in the first place to ensure their sustainable development.
All walks of life tend to see banks as institutions that can continuously issue loans, and therefore put forward various requirements and expectations for banks, such as providing support for the real economy, the rural areas, small and micro enterprises, and inclusive finance. During this process, people often neglect the fact that the source of funds to issue loans is customer deposits which have to be returned with interest in the future. This requires banks to improve not only the management of credit risk, but also the operation of assets and liabilities.
The risks facing all small and medium-sized banks in China ultimately come from these two areas as well. While commercial banks themselves should keep improving their management abilities, the society as a whole needs to bear this in mind too, instead of demanding banks offer loans with no consideration of risks.
This article was released on CF40’s WeChat blog on March 14, 2023. The views expressed herewith are the author’s own and do not represent those of CF40 or other organizations. It is translated by CF40 and has not been reviewed by the author.
