Traditional financial institutions take deposits from customers and use them to make loans. But they lend far more than they have in store at any given time – a concept known as fractional banking. On the one hand, the difference between the interest on loans and the interest paid to depositors is called the net interest margin and determines the profitability of a bank. On the other hand, the difference between assets and liabilities is called their equity and determines the resilience of the bank to external shocks.
Prior to the latest bank run, SVB was considered not only a profitable banking institution, but also a safe one, as it held $212 billion in assets against approximately $200 billion in liabilities. That means they had a cushion of $12 billion in equity or 5.6% of assets. That’s not bad, although it’s about half the 11.4% average for banks.
The problem is that recent actions by the US Federal Reserve have reduced the value of long-term debt, to which SVB was heavily exposed through its mortgage-backed securities (about $82 billion). When SVB reported to its shareholders in December that it had $15 billion in unrealized losses, wiping out the bank’s capital cushion, it raised many questions.
Related: USDC has de-indexed, but it won’t default
On March 8, SVB announced that it had sold $21 billion in liquid assets at a loss and said it would raise funds to make up for the loss. But the fact that he announced the need to raise more money – and even considered selling the bank – caused considerable investor concern, leading to an estimated $42 billion in attempted withdrawals from the bank. Of course, SVB did not have sufficient liquidity and the Federal Deposit Insurance Corporation took over on March 17.
The macro-financial literature has a lot to say about these situations, but a good summary is to to wait for highly non-linear dynamics – i.e. small changes in inputs (the equity-to-assets ratio) can have substantial changes in output (liquidity). Bank runs can be more frequent during recessions and have significant effects on overall economic activity.
Look for structural solutions
Admittedly, SVB is not the only bank to be exposed to higher and riskier macroeconomic conditions, such as interest rates and consumer demand, but that is only the tip of the iceberg. hit the headlines last week. And we’ve seen it before – most recently during the financial crisis of 2007-2008 with the collapse of Washington Mutual. The consequences led to increased financial regulation, largely in the Dodd-Frank Act, which expanded the powers of the Federal Reserve to regulate financial activity and authorized new consumer protection guidelines, including the launch of the Consumer Financial Protection Bureau.
It should be noted that the DFA also enacted the “Volcker Rule”, restricting banks to proprietary trading and other speculative investments, largely preventing banks from operating as investment banks using their own deposits to trade stocks, bonds, currencies, etc.
JUST IN: SEC launches investigation into Silicon Valley Bank $SIVB executive stock sales made days before the crash.
—Watcher.Guru (@WatcherGuru) March 14, 2023
The rise of financial regulation has led to a dramatic shift in the demand for science, technology, engineering, and math (STEM) workers, or “quants” for short. Financial services are particularly sensitive to regulatory changes, with much of the burden falling on the workforce as regulations affect their non-interest expenses. Banks realized they could reduce compliance costs and increase operational efficiency by increasing automation.
And that’s exactly what happened: the proportion of STEM workers increased by 30% between 2011 and 2017 in financial services, and much of that was attributed to increased regulation. However, small and medium-sized banks (SMEs) have had a harder time coping with these regulations, at least in part due to the cost of hiring and building sophisticated dynamic models to forecast macroeconomic conditions and balance sheets. .
The current state of the art in macroeconomic forecasting is stuck in the econometric models of 1990 which are very imprecise. While forecasts are often adjusted at the last minute to appear more accurate, the reality is that there is no single model or consensus approach to forecasting future economic conditions, setting aside some exciting and experimental approaches by , for example, the Federal Reserve of Atlanta with its GDPNow tool.
Related: Lawmakers should check SEC’s wartime consigliere with legislation
But even these “nowcasting” tools don’t incorporate large amounts of disaggregated data, making forecasts less relevant for SMEs exposed to certain asset classes or regions and less interested in the national state of the economy. economy itself.
We need to move from forecasting as a “check-off” regulatory compliance measure to being taken seriously as a strategic decision-making tool. If nowcasts aren’t working reliably, stop producing them or find a way to make them useful. The world is very dynamic and we need to use all the tools at our disposal, from disaggregated data to sophisticated machine learning tools, to help us understand the times we are going through so that we can behave prudently and avoid crises. potential.
Could better modeling have saved Silicon Valley Bank? Maybe not, but better modeling would have increased transparency and the likelihood that the right questions would be asked to prompt the right precautions. Technology is a tool—not a substitute—for good governance.
In the aftermath of the Silicon Valley Bank collapse, there was a lot of finger pointing and rehashing of the past. More importantly, we should ask ourselves: Why did the bank run happen and what can we learn?
Christos A. Makridis is a teacher and entrepreneur. He is CEO and Founder of Dainamic, a financial technology startup that uses artificial intelligence to improve forecasting, and is a research affiliate at Stanford University and the University of Nicosia, among others. He holds a Ph.D. in Economics and Management Science and Engineering from Stanford University.
This article is for general informational purposes and is not intended to be and should not be considered legal or investment advice. The views, thoughts and opinions expressed herein are those of the author alone and do not necessarily reflect or represent the views and opinions of Cointelegraph.