I am reading this book right now and goodness it is a hoot. Did you know that prior to computerization the error rates for stock trades was *high single digit percent*?
Schwab fought theirs down to *only 6%* through process improvements then bought a used computer (for ~$500k).
Schwab fought theirs down to *only 6%* through process improvements then bought a used computer (for ~$500k).
In living memory, every account statement was handcrafted by artisans. Goodness have we forgotten that.
Floors upon floors of clerks, patiently adding up numbers in a fashion to MS Excel, except with added ability to have your life savings (temporarily) disappear to math errors.
Floors upon floors of clerks, patiently adding up numbers in a fashion to MS Excel, except with added ability to have your life savings (temporarily) disappear to math errors.
Another fun takeaway in the book, which is one of the non-obvious things about the brokerage industry: Schwab was a *marketing* shop, with one extremely core product/pricing innovation that mattered in early years, but very much a marketing shop.
It dominated Schwab's concerns.
It dominated Schwab's concerns.
Schwab nearly blew through all the firm's capital buffer when someone wrote a check to a retiree for $9k (of ~1970 dollars; approximately $60k today)... except they actually wrote it for $90k.
It got deposited; took months to get back; nearly caused the firm to close.
It got deposited; took months to get back; nearly caused the firm to close.
A wild and recurring theme is that they were so undercapitalized that years into the life of the company, when they were growing like a rocket and suffering the atrocious cash flow dynamics of brokerages, they were routinely stymied by difficulty in raising $X0k of capital.
Even inflation adjusted they're talking about numbers that individual PMs at AppAmaGooBookSoft routinely angel invest an hour after meeting someone.
In a "Wow, there's a little bit of Bitcoin in everyone", apparently a senior operations person at Schwab told the SEC, which expressed concerns about Schwab's bookkeeping, "By what authority are you asking me these questions!?"
This is not the right thing to say to the SEC.
This is not the right thing to say to the SEC.
(A bit of Dangerous Professional advice: if you ever need to do that pushback to a regulatory agency, it will be delivered by your lawyer, in writing, politely. You will not usually have non-lawyers talking unsupervised to regulators to avoid pretty much exactly this happening.)
Venture capital comes up numerous times; Schwab was spurned by it, repeatedly, but apparently benefited from ambient climate in San Francisco to build something massive.
There's a line you'd swear was from a tech CEO regarding their capital intensive strategy to build branches.
There's a line you'd swear was from a tech CEO regarding their capital intensive strategy to build branches.
(Paraphrase)
"[Discount brokerage X] was happy growing slowly and didn't want to build branches, because they were capital intensive and have a 4 year payback period. And they built a nice little business for themselves. I wasn't satisfied. I built branches. We grew much faster"
"[Discount brokerage X] was happy growing slowly and didn't want to build branches, because they were capital intensive and have a 4 year payback period. And they built a nice little business for themselves. I wasn't satisfied. I built branches. We grew much faster"
Hindsight is 20/20 and etc disclaimer out of the way, two true statements:
Schwab is the largest discount brokerage in the US by a commanding margin.
You've never heard of the firm whose name I'm eliding. I do not know if it still exists.
Schwab is the largest discount brokerage in the US by a commanding margin.
You've never heard of the firm whose name I'm eliding. I do not know if it still exists.
In a line I would like to staple to the heads of all tech company marketing departments, Schwab talks about how he, as a CEO in the 1970s with a background in direct response marketing, knew his channels, costs of customer acquisition, conversion rates, and payback periods cold.
(It is broadly underappreciated in tech that e.g. A/B testing, including statistical confidence testing on it, predates computers.)
This is just amazing to me in so many ways: the desired and delivered customer experience in the early days:
Operator: "Thank you for calling Charles Schwab."
A retail customer: "My account number is [7 digits]. 2,000 shares of IBM at limit of 57. *disconnects from call*"
Operator: "Thank you for calling Charles Schwab."
A retail customer: "My account number is [7 digits]. 2,000 shares of IBM at limit of 57. *disconnects from call*"
And a few hours later Schwab would call you back:
Operator: "Filled 1,000 shares of IBM at 57; one left. Is this correct?"
Operator: "Filled 1,000 shares of IBM at 57; one left. Is this correct?"
They did material work in call centers to specialize such that e.g. only operators with broker licenses took orders, but since they were expensive, if you called and asked for a quote ("What is IBM trading at?") you got routed to effectively (cheap, tier 1) CS.
