quick follow-up on Charles Schwab

Original post: some thoughts on Charles Schwab, published Mar. 12, 2023:

In The Media Very Rarely Lies, Scott Alexander writes:

the media rarely lies explicitly and directly. Reporters rarely say specific things they know to be false. When the media misinforms people, it does so by misinterpreting things, excluding context, or signal-boosting some events while ignoring others, not by participating in some bright-line category called “misinformation”.

I was reminded of this phenomenon over the last few weeks after reading much of the commentary about Charles Schwab. Most of the long-form Schwab write-ups and Schwab-related tweets I read made the same point: marking all of Schwab’s securities to fair value wipes out nearly all of its tangible book value. This is factually true. The authors are not lying. But they are “misinterpreting things” and “excluding context”. As I hope I made clear in my last Schwab write-up, you can’t express a view on solvency without expressing a view on cash sorting. If no depositors cash sort out of the bank, then Schwab’s investment securities, predominantly Treasuries and Agencies, will pull to par as they mature and everything is fine. If all depositors cash sort at once, then Schwab Bank will be forced crystallize all its unrealized losses and find itself in ruin.

If you want to say “scuttleblurb, you’re delusional. I think Schwab will have to tap into the HTM bucket as 80% of deposits flee and (for whatever reason) liquidity backstops are unavailable, which in turn will convert huge unrealized losses into realized losses, impair their capital ratios, and either force them into receivership or into raising boatloads of capital on massively dilutive terms“, that’s great! This is the right framework for thinking about insolvency in this particular case.

Of course, I thought the chance of an SVB-style debacle was very low (say, less than 3%) at the time of my last post and I think the probability is even lower now with the Fed’s the Bank Term Funding Program (BTFP) in place. The BTFP lets Schwab and other federally insured banks to access funds against the par value of Treasury and Agency collateral. The advances will be made available for a term of one-year at a rate of ~5%1.

The day after the BTFP was announced, Schwab released its monthly Activity Highlights, which included the following statement:

We have access to significant liquidity, including an estimated $100 billion of cash flow from cash on hand, portfolio-related cash flows, and net new assets we anticipate realizing over the next twelve months. We believe we have upwards of $8 billion in potential retail CD issuances per month, plus over $300 billion of incremental capacity with the Federal Home Loan Bank (FHLB) and other short-term facilities – including the recently announced Bank Term Funding Program (BTFP).

To briefly recap, at year end Schwab’s immediate sources of funds included:

Cash: $40bn

AFS govies maturing in < 1 year: $22bn

Other AFS-securities: $123bn

On top of all that, they can now tap into $176bn of funding by posting HTM securities, all US agency MBS, to the BTFP.

That’s more than $300bn of liquidity against Schwab’s $367bn of total bank deposits, only $73bn of which is uninsured. In an interview with The Wall Street Journal last Thursday, CEO Walt Bettinger affirmed, “There would be a sufficient amount of liquidity right there to cover if 100% of our bank’s deposits ran off,”….“Without having to sell a single security.” So the risk of a catastrophic liquidity strain that tips Schwab Bank into receivership seems very very low. And in the unlikely event that half of total deposits cash sort out of the bank and Schwab is forced to liquidate all its AFS securities, its Tier 1 leverage ratio would actually improve (from 7.1% to 8.1%): the excess capital released as the balance sheet shrinks exceeds the after-tax value of realized losses on the AFS securities being sold.

But even if Schwab can survive a deposit run, it still faces an earnings challenge. The most vigorous proponents of this view are puzzled that anyone would keep brokerage cash in deposits earning ~0% when they could, with just a few clicks, invest that cash in one month T-bills earning ~4%. The pace of rate hikes is like nothing we’ve seen in the last 40 years and with broad public awareness of inflation and rates, the consumer is likely far more sensitized to rate hikes today than they were during the far more measured rate hiking cycle of 2015-2019, which management often points to as a comp for cash sorting dynamics.

Moreover, while 80% of Schwab’s deposits are FDIC-insured and come from 34mn brokerage accounts that represent a wide cross section of the US mass affluent – making them far more diversified and stable than SVB’s, which were concentrated in the accounts of cash burning startups funded by herdlike VCs with Twitter megaphones – they are also qualitatively different from primary checking accounts that consumers use for day-to-day purchases. I suspect that the brokerage account clients that own these deposits and the RIAs who manage half of Schwab’s clients’ assets are more attuned to visible and easily attainable sources of incremental yield than the average depositor at some regional bank.

On the other hand, brokerage clients will typically maintain some minimum amount of transactional cash in their account – either out of inertia or maybe just to yolo into GameStop or whatever at a moment’s notice – so cash sorting should diminish as cash balances decline. At year-end, the average Schwab brokerage client had just under ~$15k of sweep cash in their account (down from ~$18k the year before)2. That’s about 7% of their Schwab assets. Maybe clients now feel compelled to shove a big part of that residual 7% cash allocation into T-bills to juice returns. But then again maybe the average Schwab brokerage client is not as conscientious about optimizing yield as you are.

Several readers replied to my last post with something along the lines of “even if solvency risk is off the table, with rates where they are, cash sorting will be a big headwind to earnings”, as if I disagreed. I don’t. The thesis here turns on the degree and persistence of cash sorting from current levels.

(Update – 3/26: in the original post, I used changes in avg. interest-earning assets as a rough proxy for changes in deposits. This was too noisy to be useful, especially over very short time periods like 2 months, so I deleted it. Doesn’t change the analysis. I was basically trying to get at a reasonable base assumption for changes in bank deposits in 2023 in the lead up to my bear case scenario.)

Again, this doesn’t mean that cash sorting is over (no one’s saying that, not even management), just that the degree of cash sorting diminishes as rate hikes moderate. In its monthly activity statement published March 13, following a week where everyone was freaking out over SVB contagion risk, management backed up this point with some data:

Client bank sweep cash outflows in February were about $5 billion lower than January and March month-to-date daily average outflows are tracking consistent with February. Importantly, these outflows reflect a continuation of client decisions to reallocate a portion of their cash into higher yielding cash alternatives within Schwab. Based on our ongoing analysis of these trends, we still believe client cash realignment decisions will largely abate during 2023.

Over the following days, insiders bought a ton of shares:

Source: BAMSec

I can’t remember the last time I saw so much widespread insider buying in such a short window of time. Given that banking is ultimately a confidence game, it’s possible these purchases were coordinated to signal conviction even as the underlying business deteriorated far more than management anticipated, but that’s a pretty cynical read.

But one need not be cynical to be bearish. In this anxious time, hawkish Fed behavior may carry disproportionate weight given all the attention on bank balance sheets. So let’s go a little nuts with the cash sorting and say that with the Fed rate hiking by 50 bps this year3 deposits fall by 41% from year-end levels, or $150bn. That’s close to twice the net outflows that Schwab experienced last year, when rates moved from 0% to 4%+. Let’s further grant that all deposit outflows are moved into T-bills and nothing is diverted to investment vehicles where Schwab earns management fees.

In this bear case, I also assume the following:

1) the bank balance sheet never grows again

2) cash sorting outflows are funded with $40bn cash, $22bn of short-term govies, $70bn of AFS liquidations, and $18bn of BTFP/FHLB funding

3) client assets continue to grow by ~5%/year. If you’ve followed Schwab for a while you know that this company metronomically grows client assets by 5%-7%. The banking drama unfolding over the last 2 weeks hasn’t changed that. In their March 13 monthly activity statement, Schwab reported:

February core net new assets totaled $41.7 billion, our 2nd largest February ever (trailing only February 2021, the height of the meme stock craze). Our growth and momentum have continued into March, with daily net new assets averaging nearly $2B per trading day month-to-date.

They provided yet another update 4 days later:

The Charles Schwab Corporation today announced it has seen strong inflows from clients over the last week. Over the past five trading days (3/10/23-3/16/23), clients have continued to bring assets to Schwab, with approximately $16.5 billion in core net new assets for the week, demonstrating the trust clients place in Schwab.

So let’s say Schwab’s asset management fees, trading and “other” revenue grow in line with client assets, by 5%/year.

4) As a standalone company, TD Ameritrade earned “Bank Deposit Account” fees on the client brokerage cash it swept into TD Bank. Post-acquisition, these BDA balances, which totaled $127bn at year-end, will migrate to Schwab’s balance sheet at a pace of ~$10bn a year until they are reduced to $50bn. On the transferred deposits, the 1% fees it loses on BDA balances is more than made up for by the incremental profits in gains reinvesting those proceeds at higher yields. This was the biggest source of projected revenue synergies at the time of TDA’s acquisition. In this scenario, I assume BDA balances decline along with Schwab’s deposit balances, by 41% this year. And going forward I assume transferred BDA balances, instead of moving to Schwab Bank, are parked in T-Bills.

5) Last year, with the Fed raising from 0% to 4%, Schwab’s deposit rate moved from ~0% to 0.46%. I assume that this year the rate Schwab pays depositors doubles, from 0.46% to 1%.

There are several important earnings offsets to this sorry state of affairs.

First, in response to a 40% decline in net interest income that takes total revenue down 21%, I assume Schwab cuts 5% of its expense base and grows that base by 2/3 the pace of revenue growth thereafter.

Second, I estimate that Schwab still has ~$600mn of cost synergies and (conservatively) ~$600mn of non-BDA revenue synergies remaining from the TDA acquisition.

Third, even after crystallizing losses on $70bn of AFS securities to meet outflows, Schwab is more overcapitalized than it was before, with a Tier 1 Leverage ratio of 8.2%. They could liquidate and realize losses on the remaining $53bn of AFS securities, reinvest the proceeds at significantly higher yields, and still be above their year-end 7.1% ratio (and well above the 5% minimum ratio set by regulators).

Fourth, with the bank balance sheet no longer growing, Schwab doesn’t need to post incremental bank capital.

When I account for these moving parts, it looks like Schwab is doing about $5.7bn of after-tax earnings in year 5 (down considerably from its current ~$8bn run-rate). In the terminal year Schwab is growing revenue by ~2% and earnings by ~3% (more like 4% and ~5%-6%, respectively, if you look past the steady BDA fee decline). At 12x + accumulated earnings + excess bank capital4, Schwab is valued at ~$52/share in year 5, or ~$35 in present value terms, assuming an 8% discount rate (-34% from today’s price of $53). At $53 the stock appears to be pricing in ~$80bn of deposit outflows, followed perpetual bank stagnation5.

Anyway, at a high level what my bear case is saying is that Schwab emerges from this cash sorting cycle a shadow of its former self. Interest earning assets decline by 26% this year and never recover while the spread between what Schwab earns from its investments and pays to its funding sources compresses by 25%, from 2.24% in 4q22 to 1.69%.

I think the chances of this scenario playing out are pretty remote, like less than 10%. Cash sorting will eventually settle down somewhere and the enormous ~$400bn+ of net new assets that flow into Schwab every year like clockwork supplies a natural tailwind to deposit growth. Even if just 5% of net new client assets are swept into bank deposits (compared to 7% of total client assets today), that’s still an additional ~$20bn being put to work every year at juicy incremental spreads. Plus, nearly all of deposits that leave Schwab Bank still remains within the broader Schwab complex and surely some portion of that will find its way to fee-generating investment vehicles.

But my toy bear case scenario, despite its inevitable flaws, is still useful for getting a rough sense of how bad things could get, even if Schwab turns out to be fine from a liquidity/solvency standpoint. When a company might be a good investment but there is a thick fog around earnings power, establishing some semblance of a floor can give you a sense of where it might make sense to add. I have a few points of dry powder with Schwab’s name on it in case the stock trades down to like $40s (though I often change my mind and if the stock trades down to $40 for fundamental reasons that undermine my confidence, all bets are off). But I have no more Schwab appetite beyond that. In truth, as interesting as Schwab looks to me at current prices, I will never size this up to double-digits because…well…see all the caveats in my first post. Most banks are mediocre investments most of the time but if you absolutely must own one, consider John Hempton’s rule about averaging down.

Disclosure: At the time this report was posted, accounts managed by Compound Insight LLC owned shares of SCHW. This may have changed at any time since.

  1. one-year overnight index swap rate plus 10 basis points.
  2. I am including Schwab’s deposit balances + BDA balances held at TD Bank.
  3. A few days ago the Fed hiked again, by 0.25% and kept its expectation for peak Fed funds unchanged at 5%-5.25%, implying one more 0.25% hike for this year. If you’re keeping count, that’s 50 bps of implied rate hikes for 2023 compared to ~400 bps in 2022.
  4. “excess” defined as the amount of capital above Schwab’s 7.1% T1 Leverage Ratio.
  5. in this case, Schwab doesn’t need to tap high cost Fed facilities but the AFS liquidations take Schwab’s T1 Leverage Ratio below 7% but still above the regulatorily mandated 5% minimum. They can get back to 7% with less than 6 months of earnings.