Artificial intelligence is beginning to invade the institutional investment world, as it has begun to invade the industrial world.
As this invasion proceeds it likely will have a dramatic impact not only on the investment world, but also on the entire financial sector.
As reported in the April 16 issue of Pensions & Investments, some hedge funds and large investment management firms, such as BlackRock, Goldman Sachs Asset Management and J.P. Morgan Asset Management, are using machine learning and big data analysis in their investment programs.
They use powerful computers and computer models to screen vast amounts of data to find actionable investment insights far faster than humans can, and before other investors recognize and act upon them. The objective is to generate more alpha in actively managed portfolios.
In addition, the programs learn from experience. At present, the amounts of money managed using AI are small relative to the total amounts overseen by the investment giants, but that likely will change.
The full impact of those AI programs cannot be identified at present, just as their full impact on the industrial world cannot yet be seen, but all involved with the financial sector should be prepared for change. As in the industrial sector, jobs likely will change, and jobs might be lost.
For example, AI might be able to identify tiny market inefficiencies that were not detectable previously, or were so small investors could not take advantage of them because of trading costs. If those AI programs generate additional alpha, they will eventually make the stock market even more efficient than it already is, making successful active equity management even harder.
Think of the search for alpha as the search for gold. One hundred and fifty years ago, gold was found by prospectors, alone or in groups, panning in likely creeks in areas where gold was suspected. Later it was mined deep underground using power tools in areas identified by geologists. Now it is mined using huge dredges in open-cut mines as the places where the precious metal is easy to find in large amounts have been mined out.
AI is like the huge mining dredges used in gold mining today. The easy-to-recover alpha has been discovered by thousands of security analysts and recovered by successful active managers. The more powerful tools AI represents digging into big data gathered and stored in huge computers might be able to harvest the remaining fragments.
Active managers who have not invested in AI will find it harder to compete both for alpha and for clients, and many may go out of business, reducing the number of financial sector jobs for analysts and portfolio managers. AI costs lots of money, and firms using it need employees with different skills, particularly mathematical modeling and coding skills.
More managers and clients might have to adopt indexed management as active stock selection becomes less rewarding. Artificial intelligence algorithms might even discover ways to take advantage of indexing.
On the other hand, it's possible such algorithms might find subtle signals buried in big data that make active beta programs successful until the signals are arbitraged away.
AI likely also will bring changes in trading, and index arbitrage might be eliminated as AI programs learn.
AI algorithms also will affect fixed-income management as they harvest and digest deeper financial information about bond issuers, leading to more accurate bond prices.
In short, as such programs are adopted more widely, the impact on all parts of the financial sector is likely to be enormous. The leaders of those sector should be preparing their firms for the changes artificial intelligence mining big data will bring.
Ultimately, AI likely will be adopted by government departments, particularly the IRS and the SEC, leading to better enforcement of laws and regulations.n
