The 1992 book Merchants of Debt, a critical history of the private equity firm KKR, recounts the moment in the early 1980s when a firm executive came across VisiCalc, the spreadsheet software that would upend both KKR and Wall Street.
“KKR couldn’t rapidly stalk several companies at once, because its financial blueprints required weeks of calculations by hand,” George Anders wrote of the old way of doing business. But with the arrival of VisiCalc “all of a sudden, giant companies’ finances could be picked apart in an afternoon”.
After VisiCalc came Lotus 1-2-3 and later Microsoft Excel. For investors like Martin Brand, who today runs Blackstone’s US leveraged buyout group, it is now possible to get a deal analysis done while still on the phone to the investment banker making the sales pitch.
Blackstone has dozens of data scientists and engineers. They work alongside its investors to come up with automation tools able to ease the processing burden on junior employees without sacrificing the rigorous number-crunching needed to put billions of dollars to work.
One such tool is BX Atlas, a standardised LBO model that gives Blackstone dealmakers a near-instantaneous readout of a deal’s feasibility and whether it merits a more detailed study. “It’s really cool,” said Brand.
Even with functionality increasing each year, 20-something bankers and investors had been stuck manually typing numbers and equations into cells and then linking and formatting those cells.
That drudgery may be about to end with seismic advances in computing power. What that means for investment returns, the happiness of workers in the salt mines of Wall Street and the skills needed for the next generation of rainmakers has become the next interesting question.
For firms lacking Blackstone’s internal tech developers, Ian Gutwinski is trying to fill the gap. The 2021 Harvard Business School grad and former PE investor sells LBO modelling software called Mosaic. Like Blackstone’s Atlas, Mosaic can spit out investment returns with just a few quick manual inputs.
Mosaic also delivers helpful visualisations that show whether those returns stem from operating improvements or sheer financial leverage. The whole process takes minutes. Gutwinski says that preliminary models are not only fast but defect-free, eliminating the chance of common human error such as entering the wrong number or link. Mosaic models can also be downloaded to Excel and then further customised as deemed necessary by the deal team.
Gutwinski’s clients so far include powerhouses such as Warburg Pincus and CVC Capital. One user, who declined to be identified, explained that junior associates had numerous daily responsibilities, including monitoring existing companies, that were better uses of their time. “Modelling is a commodity,” this person said.
“One of the really striking things about the story of spreadsheets in the 1980s, particularly in the case of PE, is that spreadsheets allowed financiers to see and imagine differently,” said William Deringer, a historian at MIT and a former investment banker.
“I think it is quite possible that new automation tools for financial analysis might offer similarly new kinds of vision and imagination.”
In an academic paper, Deringer cited a 1989 New York Times article that captured the cultural shift sweeping Wall Street in the nascent VisiCalc era. “Sharp elbows and a working knowledge of computer spreadsheets suddenly counted more than a nose for dry sherry or membership in Skull and Bones”, it wrote — the latter referring to Yale’s blue-blooded secret society.
At this time of year, investment banks are greeting their new trainee classes and teaching them the building blocks of spreadsheet modelling. The ultimate objective is not just understanding the mechanics but building critical thinking, intuition and problem-solving abilities. And the coin of the realm in these fresh cohorts historically has been slick Excel skills.
It is possible that emerging black-box tools will soon mean that technical capabilities matter less, in the same way that software coders are expected to one day give way to AI “prompt” engineers. There has always been a gap between what makes a good junior associate — raw processing horsepower — and a senior executive — decisiveness, maturity, and a respectable golf game. Starting now, that disconnect may slowly begin to shrink.