Markowitz chose to apply mathematics to the analysis of the stock market as the topic for his dissertation. Jacob Marschak, who was the thesis advisor, encouraged him to pursue the topic, noting that it had also been a favorite interest of Alfred Cowles, the founder of the Cowles Commission. While researching the then current understanding of stock prices, which at the time consisted in the present value model of John Burr Williams, Markowitz realized that the theory lacks an analysis of the impact of risk. This insight led to the development of his seminal theory of portfolio allocation under uncertainty, published in 1952 by the Journal of Finance.
In 1952, Harry Markowitz went to work for the RAND Corporation, where he met George Dantzig. With Dantzig's help, Markowitz continued to research optimization techniques, further developing the critical line algorithm for the identification of the optimal mean-variance portfolios, relying on what was later named the Markowitz frontier. In 1954, he received a PhD in Economics from the University of Chicago with a thesis on the portfolio theory. The topic was so novel that, while Markowitz was defending his dissertation, Milton Friedman argued his contribution was not economics. During 1955–1956 Markowitz spent a year at the Cowles Foundation, which had moved to Yale University, at the invitation of James Tobin. He published the critical line algorithm in a 1956 paper and used this time at the foundation to write a book on portfolio allocation which was published in 1959.