• A “one-two punch” for the banking industry
  • Can GPT-4 make us millionaires?
  • CBDC’s and bank runs…

Dear Reader,

Welcome to our weekly mailbag edition of The Bleeding Edge. All week, you submitted your questions about the biggest trends in tech and biotech. Today, I’ll do my best to answer them.

If you have a question you’d like answered next week, be sure you submit it right here. I always enjoy hearing from you.

“Complete and utter malfeasance” at SVB…

Hi Jeff,

I have been an Unlimited Member for quite a few years, always enjoy your well-written and easy-to-read commentary and analysis, especially on technology trends. I was involved with microprocessor development in the very early years (70s decade) at General Instruments architecting some of the chips that became the backbone of the early days of Microchip Technology in Chandler, AZ.

Of course your explanation about what happened at SVB was accurate, but blaming the Fed for their problems is just really a red herring – more spin to put the blame on someone or something else other than the real culprit.

Yes, the FED raised rates a lot over the course of 2022, but they didn’t do it in secret. Jay Powell and one or more of a host of other senior Fed executives speak publicly somewhere or other almost every week. They publish notes of their bi-monthly meetings, unlike all corporations that conduct internal business strategy meetings in complete secrecy (which is why they are so inflamed when there are leaks).

The real problem at SVB was not the Fed, it was the complete and utter malfeasance in office (really negligence) of SVB management and its Board of Directors.


All these people are supposed to be competent, experienced, business management executives. They are not living on another planet, they all were well aware that the Fed was raising rates rapidly in 2022, it was no secret. They all were aware that the bank’s deposit base was not hundreds of thousands of everyday people with less than $250K in their accounts, but rather concentrated in a relatively small number of VC financed customers with large balances – over 94% of dollar deposits were above the $250K insured limit.

The CEO, CFO, and management team were, in theory, experienced running a bank, they were not beginners.

The CRO is responsible for evaluating the institution’s balance sheet strength and ongoing risks and challenging the CEO/CFO on their investment strategies. Where were the Board’s investment committee and risk management committee while these investment decisions were made all throughout 2022 as the rates were climbing?

Obviously not doing their jobs.

Jeff S.

Hi, Jeff. Thanks for writing in with your views on this incredible situation. We have a common corporate background as well. GI was the first tech company that I worked for after getting my engineering degree. That was in the 90’s, a bit later than you, but at least it was still known as GI before the name change and the ultimate breakup and sale.

If you were able to read Wednesday’s Bleeding Edge this week, you’ll know that I agree and share your concerns with the management of the bank. 

There is a long list of things that I don’t like about what was going on at SVB. It got me thinking about whether or not SVB and Signature were one-offs, or if there was greater concern of something far more widespread.

We know that the Fed’s rate hiking regime has been the fastest in recent history, as the chart below shows.

After more than twenty years of low rates, the Fed hiked rates by 475 basis points in just a few months to where we are today, at 5%.

Even if SVB had been more diligent with its duration risk, there’s no way the entire banking sector could have pivoted quick enough to eliminate all risk. The Fed always knew its rate hikes were going to “break” something.

And we don’t even have to speculate on that. Here’s a few past examples:

  • When the Fed got aggressive in 1999/2000 it drove the Fed Funds rate to 6.5% and broke the NASDAQ

  • In 2005 – 2007 it hiked interest rates 425 basis points to 5.25% and it broke the housing market

  • Starting in 2016 through 2018, it hiked interest rates 225 basis points until the repo markets broke, which resulted in a crash of the stock markets

And now the Fed has hiked farther and faster than almost ever before. Something, or things, are starting to break.

As I wrote on Wednesday, The Kansas City Fed admitted that the rising rates were creating unrealized losses in the banks’ AFS bond portfolios. AFS stands for “available for sale” and they’re a critical tool for banks to raise liquidity at times like these.

The Kansas City Fed wrote:

At year-end 2021, only 4 community banks had tangible equity capital ratios below 5 percent; that number increased to 333 at June 30, 2022, indicating less ability to sustain economic shocks.

The Fed knew all the way back in June that its policies were putting a strain on the banks, and they continued hiking rates all the same.

And the reality is that the only reason these banks had such a massive bond portfolio in the first place is because of government policy.

During 2020 and 2021, these banks saw a massive influx of new deposits as a direct result of stimulus spending. The New York Times actually covered this back in October of 2021 in an article titled “Banks are Binging on Bonds, but Not Because They Want to.”

Much of that money has poured into the bank accounts of American households and companies. By the end of May, nearly $830 billion in stimulus check payments had been sent to individuals.

Roughly $800 billion more was sent to businesses in the form of programs such as the Paycheck Protection Program. And about $570 billion was spent on extended and enhanced unemployment insurance benefits, according to data from the Government Accountability Office.

Under more normal circumstances, banks would make loans with these new deposits. But the influx was so large and so fast that loan making simply couldn’t keep up. Banks were pushed into government debt. The banks bought $150 billion worth of Treasurys in the second quarter of 2021 alone.

In essence, fiscal policy flooded banks with artificially high levels of new deposits, forcing them into Treasurys. Then, monetary policy destroyed the value of those bonds in record time. It was a one-two punch.

Could SVB have done things differently? Absolutely. But I’m not so willing to let the Fed—and the government, for that matter—off the hook so easily.

And without the recently announced Bank Term Funding Program (BTFP), we would have already seen the collapse of a number of other regional banks for exactly the same reason.

The real question that we need to be asking is ‘what else is going to break?’

Can GPT-4 make us money?


I really like your content and calm clean style.

Question: if AI is so fast and smart and has almost all human data at its disposal, can’t I just ask it to turn $1K into $1.5M for me?

Richard L.

Hi, Richard. What a fun question. And you’re not the only person asking it. In fact, people are already experimenting with this idea.

Recently, a human gave GPT-4 a task: Starting with $100, make as much money as you can as fast as possible. Here’s the full prompt below:

GPT-4 Is Tasked With Making Money

Source: Twitter

And as it happens, “HustleGPT” did have a business plan in mind. The AI devised a plan to create a website to share ideas and recommendations around eco-friendly products and services. The AI chose a domain name for the website and even designed a logo using text-to-image generator Dall-E. The AI also wrote the first article published on the site.

Within a few days, the company had investors and was valued at $25,000. The company also had $1,378 on its balance sheet.

Not too bad for a start.

So to your point, could a powerful generative AI be able to build on that, scale up, and build a million dollar windfall? After a period of time, to a sophisticated user, the answer is likely yes.

Let’s take the financial markets for example. Given a large enough data set to train up on, and provided access to real time information, a powerful AI could generate outsized market returns in a specific asset class. There would have to be some human involvement to “manage” the AI; but I believe it can be done.

My team and I have used similar techniques with a powerful AI built on neural network technology to provide outsized trading returns in the cryptocurrency markets. We couldn’t have launched the technology at a worse time (March 2022) when the digital asset markets were collapsing. And yet we’ve had almost a 90% win rate on all closed trades to date.

This kind of technology will be used more and more in financial markets, and it will be used to do the kind of “heavy lifting” in the form of massive amounts of data analysis that is nearly impossible for us humans to manage.

I think a more realistic framework that we can use to understand how generative AIs can help grow wealth and make money is to view this technology as a tool, an accelerant to growth and new ideas.

We could ask it to perform a competitive analysis on a specific market and make suggestions about where the most profitable business opportunity exists in that segment.

We can task an AI to develop three versions of marketing copy that should perform well in the target demographic segments for a product.

It can be used to design a product that doesn’t exist, but could become a smash hit.

I’m sure we can think of a long list of possible scenarios where the AI can add incredible value to business and wealth creation.

The tricky variable is that nearly everyone will have access to this kind of technology. And once the whole market starts using the technology, the arbitrage opportunities quickly dissipate.

We’re in an incredible window of opportunity right now that will last a year or two. It’s the window where early adopters of this technology, or investors in this technology, will have a marked advantage over those that aren’t using it or investing in it.

The best thing that we can do is to educate ourselves on the technology, experiment with it, apply it, and if possibley gain investment exposure to it either through our own companies or companies that we invest in.

Could a CBDC prevent a bank run?

Hi Jeff,

If companies (and people) had their cash stored in CBDC wallets on a (government controlled) chain rather than bank accounts, wouldn’t that prevent a bank run, as we witnessed this week?

Tomar K.

Hi, Tomar. It’s a great question, and you must have been reading my mind. I wrote about this same topic in yesterday’s Bleeding Edge.

The answer is probably yes.

Yesterday, I predicted that the forthcoming FedNow program, which enables real-time transfers of digital dollars between financial institutions, could be enabled to act in a way similar to the Bank Term Funding Program (BTFP) that was already announced to avoid a collapse of the regional banking industry.

My thoughts were more about how this kind of technology would be used at an institutional level to avoid bank runs. But it relates directly to a digital dollar wallet or CBDC digital wallet. After all, CBDC digital wallets will be managed – custodied – by Federal Reserve-approved financial institutions.

Because of this, a CBDC wallet would prevent bank runs, at least how we understand them today. In fact, I wouldn’t at all be surprised to hear this as an argument from policymakers to “sell” this idea to the public.

At its most basic level, a bank run is a psychological phenomenon. Because of fractional reserve banking, these institutions do not have 100% liquid reserves to back up all deposits. Not even close. Just try withdrawing $50,000 from your local bank and see how long it takes to get your money.

Most people know this. If every single depositor lined up to get their money out, there simply wouldn’t be enough cash to make every depositor whole.

If people begin to lose faith in the bank, it becomes a prisoner’s dilemma. If everybody stays calm, the bank will stay solvent, and everybody remains whole.

But if people begin to “defect” (begin withdrawing their money) then the last people in line will be left with nothing. It’s this fear of being last to withdraw your funds that sparks a bank run.

But a digital wallet is fundamentally different. It’s not a bank with only a fraction of deposits kept in reserves. Instead, it’s exactly what it sounds like. It’s simply a wallet that holds a digital currency. And nobody fears that there would be a “run” on their physical wallets or the valuables kept in their personal safes.

But that doesn’t mean that a CBDC couldn’t spark a different type of panic. Let’s use our imagination a little bit…

One of the benefits of a CBDC (to central banks and policymakers, anyway) is that it could “nudge” people to make the “correct” decisions with their money.

Is the government interested in addressing climate change? It could simply limit the amount of gasoline or meat that we purchase. Or perhaps the Fed is interested in stimulating consumer spending. The central bank could put a “use it or lose it” timer on funds deposited into our digital wallets.

How would people respond to policies like this?

The reality is that we just don’t know. After all, nothing like this has been attempted before.

But it is possible that people may realize that their money that they earned is not really theirs to use, spend, and save as they see fit.

A sudden realization like this could come as a shock. Perhaps it would prompt people to exchange their CBDCs quickly for items like precious metals, other hard assets, or digital assets like bitcoin. It’s not out of the question.

The U.S. dollar is the world’s reserve currency. And a digital version of the dollar would retain that title. But what if citizens simply don’t want to hold these currencies because of onerous restrictions?

What would that do to the financial system and the strength of U.S. currency?

Like I said, we simply don’t know the answers. But I suspect we’ll find out sooner than we think….


Jeff Brown
Editor, The Bleeding Edge