CASE STUDY: SILICON VALLEY BANK – THE EARLIER YEARS
Deception And Truth Analysis (D.A.T.A.), Inc
Highly accurate and rapid deception & truth analysis of documents, creating alpha & saving due-diligence research time.
Earlier in 2023 Deception And Truth Analysis published a deep dive case study of Silicon Valley Bank. In our study we demonstrated our algorithm’s ability to identify the key issues at the bank well in advance of their meltdown. In our initial analysis, given time constraints, we were not able to trace a history of these deceptive issues back to their origin point.
Now with the luxury of additional time we can trace back the key issues to the earlier years, specifically to the first quarter of 2018. This is a full 4.5 years of advance warning of a ticking time bomb. This lead time afforded auditors, investors, and regulators in which they may have been able to assist the bank in taking corrective actions.
Unlike our previous detailed analysis, this is more a high-level overview. For those interested in a bit more detail see the Appendix to this article. There, we provide details about the contents of SIVB’s most deceptive paragraph in each 10(k) or 10(q) dating back to the first quarter of 2018. But wait, there’s more! We are premiering a new DATA functionality in the Appendix whereby we highlight not just the most deceptive or truthful sections of documents, but also the most deceptive or truthful paragraph within a section, too.
Conclusion
As you can see from the excerpts from Silicon Valley Bank’s regulatory filings in the Appendix, below, dating back to the first quarter of 2018, the company was deceptive about the composition of its loan portfolio, its net interest margin, its credit quality indicators, and its noninterest income.
There is only one quarter during the period 1Q 2018 thru 4Q 2022 in which these key issues are not the most deceptive sections of their regulatory filings. For the one quarter this is an exception – 2Q 2021 – the key issues that led to SIVB’s meltdown were what featured in their second most deceptive fragment.
For Deception And Truth Analysis this is yet another case study in which we have demonstrated our ability to identify key company issues years before markets, regulators, and auditors. Equipped with this information, our Clients can: confirm or deny critical parts of their investment theses; ask deeper questions about key deceptive issues to ensure that they have properly assessed them; adjust their assumptions in their financial models; and so on.
Appendix: Detailed Breakdown of SIVB 10(k)s and 10(q)s 2018 through 2022
2022 10(k)
DATA Score = -54.197%, 99.9986thpercentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “The embedded rate floors are also a factor in the up-rate scenarios to the extent a simulated increase in rates is needed before floored rates are cleared. In the upward parallel simulated rate shock scenarios, interest income on assets that are tied to variable rate indexes, primarily our variable rate loans, are expected to benefit our base 12-month NII projections. The opposite is true for downward rate shock scenarios.”
2022 10(q), third quarter
DATA Score = -60.909%, 99.9999thpercentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “A reduction of credit losses of $29 million for the three months ended September 30, 2021 was driven primarily by a reduction in credit losses due to the previously discussed model enhancements (see “Results of Operations—Provision for Credit Losses— (Reduction) Provision for Loans.”), a decrease in net new nonaccrual loans and recoveries, partially offset by an increase in charge-offs not specifically reserved for at June 30, 2021 and an increase in provision for loan growth.”
2022 10(q), second quarter
DATA Score = -76.172%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Noninterest income increased by $88 million for the three months ended June 30, 2022 related primarily to an overall increase in our non-GAAP core fee income. The overall increase was due primarily to higher client investment fees driven by improved fee margins resulting from higher short-term interest rates driven by the 2022 Federal Funds rate hikes.”
2022 10(q), first quarter
DATA Score = -52.014%; 99.9971st percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Interest rate risk is managed primarily through strategies involving our fixed income securities portfolio, available funding channels and capital market activities. In addition, our policies permit the use of off-balance sheet derivatives, such as interest rate swaps, to assist with managing interest rate risk.”
2021 10(k)
DATA Score = -51.764%, 99.9969th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Noninterest expense increased $166 million to $212 million in 2021 related primarily to compensation and benefits expense. Compensation and benefits expense increased as a result of an increase in average number of FTEs due primarily to the acquisition of Boston Private and higher incentive compensation and salaries and wages expenses primarily as a result of strong performance during 2021.”
2021 10(q), third quarter
DATA Score = -66.525%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “A $191 million increase in interest income from our fixed income investment securities due primarily to an increase of $61.2 billion in average fixed income investment securities driven by exceptional average deposit growth. The increase in interest income from growth of our average fixed income investment securities was partially offset by declines in yields earned on these investments reflective of the lower interest rate market environment, and…”
2021 10(q), second quarter
DATA Score = -69.539%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
[Note: SIVB’s second most deceptive paragraph for this quarter was its description of net interest income, credit losses, et al.]
Most deceptive paragraph: “Deposit service charges were $28 million and $53 million for the three and six months ended June 30, 2021 compared to $21 million and $45 million for the comparable 2020 periods. Deposit service charges increased due to the increases in product revenues from strong deposit growth and higher transaction volumes.”
2021 10(q), first quarter
DATA Score = -62.185%, 99.999th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Net interest income from GCB increased by $154.8 million for the three months ended March 31, 2021, due primarily to an increase in loan interest income resulting primarily from higher average loan balances, partially offset by a decrease in loan yields as a result of the shift in the loan mix.”
2020 10(k)
DATA Score = -59.27%, 99.9997th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “For each individual client, we establish an internal credit risk rating for that loan, which is used for assessing and monitoring credit risk as well as performance of the loan and the overall portfolio. Our internal credit risk ratings are also used to summarize the risk of loss due to failure by an individual borrower to repay the loan. For our internal credit risk ratings, each individual loan is given a risk rating of 1 through 10. Loans risk-rated 1 through 4 are performing loans and translate to an internal rating of “Pass,” with loans risk-rated 1 being cash secured. Loans risk-rated 5 through 7 are performing loans; however, we consider them as demonstrating higher risk, which requires more frequent review of the individual exposures; these translate to an internal rating of “Criticized.” All of our nonaccrual loans are risk-rated 8 or 9 and are classified under the nonperforming category. Loans rated 10 are charged-off and are not included as part of our loan portfolio balance. We review our credit quality indicators on a quarterly basis for performance and appropriateness of risk ratings as part of our evaluation process for our allowance for credit losses for loans.”
领英推荐
2020 10(q), third quarter
DATA Score = -66.521%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Noninterest expense increased by $44.2 million for the three months ended September 30, 2020, due primarily to compensation and benefits expense and professional services expense, partially offset by a decrease in business development and travel expense. Compensation and benefits expense increased $41.5 million as a result of higher salaries and wages expenses and higher incentive compensation expense. The increase in GCB salaries and wages was due primarily to an increase in the average number of FTEs at GCB, which increased to 3,102 FTEs for the three months ended September 30, 2020, from 2,364 FTEs for the comparable 2019 period. Incentive compensation expense increased due primarily to an increase in our 2020 full-year projected financial performance. Professional services expense increased due to higher expenses primarily related to our continued effort towards investments in our infrastructure, initiatives and operating projects to support our presence both domestically and globally. Business development and travel expense decreased primarily due to the impact of COVID-19 on the global economy and our restrictions placed on domestic and international travel beginning March 2020.”
2020 10(q), second quarter
DATA Score = -62.115%, -99.9999th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “For each individual client, we establish an internal credit risk rating for that loan, which is used for assessing and monitoring credit risk as well as performance of the loan and the overall portfolio. Our internal credit risk ratings are also used to summarize the risk of loss due to failure by an individual borrower to repay the loan. For our internal credit risk ratings, each individual loan is given a risk rating of 1 through 10. Loans risk-rated 1 through 4 are performing loans and translate to an internal rating of “Pass,” with loans risk-rated 1 being cash secured. Loans risk-rated 5 through 7 are performing loans; however, we consider them as demonstrating higher risk, which requires more frequent review of the individual exposures; these translate to an internal rating of “Criticized.” All of our nonaccrual loans are risk-rated 8 or 9 and are classified under the nonperforming category. Loans rated 10 are charged-off and are not included as part of our loan portfolio balance. We review our credit quality indicators on a quarterly basis for performance and appropriateness of risk ratings as part of our evaluation process for our allowance for credit losses for loans.”
2020 10(q), first quarter
DATA Score = -57.153%, 99.9995th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “…$22.4 million driven by an increase in our expected credit loss for our Investor Dependent loan portfolio given the higher relative risk and longer-duration, which is taken into account under the CECL methodology, partially offset by a decrease for our Private Equity/Venture Capital loan portfolio, given its higher historical credit quality and shorter duration. The remaining $231.6 million increase was due primarily to an increase of $190.7 million related to the expected credit losses for our performing loan reserves based on our forecast models of the current economic environment and $40.9 million related to period-end loan growth of $2.8 billion.”
2019 10(k)
DATA Score = -61.466%, 99.9999th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “The growth of $12.4 billion in total deposits consisted of $1.7 billion in noninterest bearing deposits and $10.7 billion in interest-bearing deposits. The significantly larger increase in the growth of our interest-bearing deposits relative to our loan growth, in conjunction with the increase in rates paid on our interest-bearing deposits during 2019, results in a downward impact on our EVE in the +100 bps and +200 bps rate shock scenarios. As of December 31, 2019, interest-bearing deposits represented 63 percent of total loan balances compared to 36 percent at December 31, 2018. In addition, purchases in fixed income investments during 2019 consisted primarily of mortgage-backed securities which have lower prepayment rates during increasing interest rate environments also resulting in a downward impact on EVE in the +100 bps and +200 bps rate shock scenarios reflective of lower reinvestment opportunities in an increasing interest rate environment. The downward EVE impacts in the +100 bps and +200 bps rate shock scenarios from the interest-bearing deposit growth and increase in the mortgage-backed securities are partially offset by the upward EVE impact from loan growth.”
2019 10(q), third quarter
DATA Score = -72.762%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “The provision for loan losses of $36.0 million for the three months ended September 30, 2019 reflects an increase of $19.1 million for net new nonaccrual loans, $18.3 million for charge-offs not specifically reserved for and $15.2 million in additional reserves for period-end loan growth, partially offset by a decrease of $13.0 million for the qualitative component of our performing loans and $3.9 million of recoveries.”
2019 10(q), second quarter
DATA Score = -71.635%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “For the three and six months ended June 30, 2019, noninterest income was $333.8 million and $614.1 million, respectively, compared to $192.7 million and $348.2 million for the comparable 2018 periods. For the three and six months ended June 30, 2019, non-GAAP noninterest income, net of noncontrolling interests was $315.0 million and $592.1 million, respectively, compared to $183.2 million and $325.7 million for the comparable 2018 periods. For the three and six months ended June 30, 2019, non-GAAP core fee income including investment banking revenue and commissions was $220.5 million and $438.6 million, respectively, compared to $123.1 million and $238.1 million for the comparable 2018 periods. For the three and six months ended June 30, 2019, non-GAAP core fee income was $157.3 million and $311.6 million, respectively, compared to $123.1 million and $238.1 million for the comparable 2018 periods. (See reconciliations of non-GAAP measures used below under “Use of Non-GAAP Financial Measures”.)”
2019 10(q), first quarter
DATA Score = -59.252%, 99.9997th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “The following table summarizes our impaired loans as they relate to our allowance for loan losses, broken out by portfolio segment and class of financing receivable as of March 31, 2019 and December 31, 2018:”
2018 10(k)
DATA Score = -71.272%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “A $13.7 million increase in interest income from our Federal Reserve deposits to $35.2 million, compared to $21.5 million in 2017. The increase was due primarily to higher yields as a result of rate increases in 2018.”
2018 10(q), third quarter
DATA Score = -54.521%, 99.9987th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “Noninterest expense increased by $2.1 million for the three months ended September 30, 2018, due primarily to increased compensation and benefits expense. Compensation and benefits expense increased as a result of increased salaries and wages, reflective of the increase in number of average FTE since the third quarter of 2017, and higher incentive compensation expenses, reflective primarily of our strong 2018 full-year expected performance as compared to 2017.”
2018 10(q), second quarter
DATA Score = -68.528%, 100th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “A $149.8 million increase in interest income on loans to $627.4 million for the six months ended June 30, 2018, compared to $477.5 million for the comparable 2017 period. The increase was reflective of an increase in average loan balances of $4.0 billion and an increase in the overall loan yield of 45 basis points to 5.20 percent from 4.75 percent. Gross loan yields, excluding loan interest recoveries and loan fees, increased to 4.64 percent from 4.12 percent, reflective of the benefit of interest rate increases, partially offset by the strong growth of our lower yielding private equity/venture capital loan portfolio. Our private equity/venture capital loan portfolio represented 46.8 percent and 42.2 percent of our total gross loan portfolio at June 30, 2018 and 2017, respectively.The increase in gross loan yields of 52 basis points were partially offset by a decrease in fee income from early pay-offs reflective of the increasing rate environment,”
2018 10(q), first quarter
DATA Score = -53.999%, 99.9985th percentile deceptive
Themes discussed in its most deceptive section include:
Most deceptive paragraph: “The following table summarizes our impaired loans as they relate to our allowance for loan losses, broken out by portfolio segment and class of financing receivable as of March 31, 2018 and December 31, 2017:”