Modeled on: Financial analyst, investor and professor at Columbia · founder of value investing (1894–1976)
“Buy well below intrinsic value, with a margin of safety.” Businesses priced well below a conservative estimate of their asset or earnings value, with a margin of safety.
Not affiliated with Benjamin Graham or Graham-Newman Corporation. This bot is a computer model Pick100 built from Benjamin Graham's public statements and filings. Benjamin Graham did not create, review or endorse it, its picks are not Benjamin Graham's picks, and chatting with it is not talking to Benjamin Graham. Legal notice
01The investor's record
These are the investor's own results as of 2026-10-22, not the bot's. The bot is an independent model, not endorsed by Benjamin Graham or Graham-Newman Corporation, and not investment advice.
02History and philosophy
Graham went to Wall Street after Columbia, ran his own firm from January 1, 1926 with $450,000, and with Jerome Newman ran Graham-Newman from 1936 to 1956. He taught at Columbia for many years, wrote Security Analysis (1934, with David Dodd) and The Intelligent Investor, and counted Warren Buffett among his students. Wikipedia puts Graham-Newman's return at about 20% a year against 12.2% for the market over 1936 to 1956. He died on September 21, 1976, in Aix-en-Provence, France.
His method was to buy a share as a piece of a business, only when the price sat well below a conservative value drawn from assets, average earnings and dividends. He treated the market as a moody partner (Mr. Market) and called the margin of safety the central idea. The 1929 to 1932 crash taught him a horror of borrowed money. By 1976 he had stopped believing in elaborate analysis and preferred simple price rules applied to a whole group of stocks.
03Public positions
Shares at $68.88 while net assets were worth at least $76.23: an early asset-value bargain.
Bought Du Pont, sold GM short against it, and closed the short at no loss.
Stock at $65 against $90 a share of securities; a proxy fight forced a cash payout.
The fund fell about 20%, 50%, 16% and 2% in four years. He drew a horror of borrowed money.
About 20% a year against 12.2% for the market, from net-current-asset bargains, arbitrage and hedges.
Half the company for $712,500, about 20% of the fund against a usual cap near 5%. It rose more than 200-fold.
He judged the market level of most growth stocks too high for an adequate margin of safety.
Buy below net current assets, hold 30 or more, sell at a 50% to 100% gain or after two to three years.
04How the bot thinks
Every Nasdaq 100 company is ranked on what the playbook cares about. Its long score, for how well a company fits:
The closest fits to the checklist: cheap on book value, cheap on earnings, little debt. Score near 100.
A separate short score hunts the worst fits: expensive on book value, expensive on earnings, low dividend. Re-ranked every quarter, at the close. Loss-makers are never long.
How far a company sits from the checklist: expensive on book value, expensive on earnings, low dividend.
05Investment themes
"The margin of safety is always dependent on price paid." Its job is "rendering unnecessary an accurate estimate of the future," and it goes with wide diversification.
Buffett retells Graham's picture of quotations from an obliging partner whose moods swing. "Mr. Market is there to serve you, not to guide you."
The defensive investor wants size, a strong balance sheet, ten years of earnings, a long dividend record and moderate P/E and P/B. In 1976 he simplified to earnings yield at least twice the AAA bond yield.
He wrote it and then said it was not recommended for investors. Growth projections are unreliable, and he held most growth stocks too dear for a margin of safety.
For the individual: "at least 25 per cent of the total at all times in each category," with "a consistent 50-50 division" a good default (1976).
06Model output
Long: #1 of 100 on Benjamin Graham's long-minus-short score (0.578).
Short: #1 of 50 shorts on Benjamin Graham's long-minus-short score (-0.694).
07The bot's own results
6 of 8 right
The side its portfolio held on each day's stock. The bot doesn't pick: it holds all 100.
08The full book
Equal weight: $1,000 a position. Cheap and Quality are percentiles within the Nasdaq 100; a dash means the figure is missing.
09Go deeper
10Method and limits
The 100 Nasdaq 100 companies. Anything outside it can't be held.
Long score: Book yield 3, Earnings yield 3, Leverage 3, Dividend yield 2, Net margin 1, Volatility 1, Beta 1. Short score: Book yield 3, Earnings yield 3, Dividend yield 2, Leverage 2, Net margin 1. Each a weighted blend of the company's percentile on each measure; the 50 with the best long minus short go long.
Always 50 long and 50 short. No market or rate bets. Changes fill at the next close.
Some tickers are missing figures and sit mid-pack on that measure.
Modeled on Benjamin Graham's published approach. It is not Benjamin Graham, and its picks are not Benjamin Graham's. A free game on real closing prices, not investment advice.