How Robo-Advisors Build Portfolios Using Optimization and Tilts

# Which algorithms do robo-advisors use?


# Which algorithms do robo-advisors use?












Some pundits claim that there is a revolution in portfolio management under way: The rise of the robots, a.k.a. robo-advisors. The most well known are Betterment.com, FutureAdvisor, Schwab Intelligent Portfolios and Wealthfront.

According to wikipedia

> robo-advisors employ algorithms such as Modern portfolio theory that originally served the traditional advisory community, which has used algorithmically-based automated investment solutions (dubbed in the industry as "rebalancing software") to conduct portfolio management.

My question Do you know whether there is some more information available which algorithms these firms use exactly? Is it just good old MPT or also more sophisticated stuff? (In this context it is interesting that they often get totally different portfolios). Best would be some kind of overview of the algorithms used by different providers.

## Answer by vonjd (score 12, accepted)

https://quant.stackexchange.com/a/21405

After having done a lot of research on the topic I found the following excellent research piece on ETF.com:

> Wealthfront modifies historic asset-class returns with current market implied expected returns (Black-Litterman) as well as with the in-house views of Chief Investment Officer Burton Malkiel’s team. In addition, Wealthfront sets minimum and maximum weights for each asset type. The resulting portfolio has an unmistakable Malkiel flavor to it, with an emerging market allocation that reflects his interest in China. Betterment uses Black-Litterman currently implied market expected returns, but deliberately includes small-cap and value as separate asset classes, adding a classic Fama-French factor tilt. It doesn’t constrain the portfolio weights, but they do account for downside risk. Betterment’s portfolios wind up quite similar to the global market, at least on the equities side. Covestor deliberately veers away from its optimizer to hedge its portfolios against inflation and to adjust for downside risk. Its wide constraints allow heavy weights to emerging markets. Wise Banyan constrains its portfolio weights “tighter than most,” back toward market-cap weights, according to Herbert Moore, co-founder and chief investment officer. This might explain why its portfolios allocate generously to U.S. equities, and away from the rest of the global equity market. Invessence includes the largest number of asset types, adding granularity to the fixed-income side. It bases asset-class returns expectations on up to 80 years of historical ETF or index returns, but uses only nine years of volatility history. Invessence employs gold as an inflation hedge. It also constrains all asset weights except for U.S. equity. Sure enough, the U.S. dominates its equity allocation. FutureAdvisor doesn’t optimize. Instead, its builds its portfolio in sleeves, creating a glide path much as the target-date mutual funds do. It builds in a “strategic” allocation to REITs as an inflation hedge, adding Fama-French type tilts. They’re not kidding. The firm’s portfolios emphasize small- and midcap stocks, and financials (REITS), with highest-in-class dividend yields and lowest price/book ratios.

There a many more details here: http://www.etf.com/sections/blog/22973-ghosts-in-the-robo-advisor-machine.html?nopaging=1 and here: http://www.etf.com/sections/blog/22982-inside-robo-advisor-asset-allocation.html?nopaging=1

The whole 7-parts series on the topic starts here: http://www.etf.com/sections/blog/22946-which-robo-advisor-for-my-teen.html?nopaging=1

## Answer by RockZen (score 5)

https://quant.stackexchange.com/a/30441

Well, I did some modest research on this topic, looking at peers. Most of them use Modern Portfolio Theory, see this pic:

You can find this small survey here: https://www.linkedin.com/pulse/roboadvisors-like-commodore-vic20-apparently-according-raffaele-zenti?trk=mp-reader-card

The sector, I mean Roboadvisors, has a lot of disruption potential, obviously. But it is still immature, methodologically, with respect to the "traditional" (i.e. offline) asset management/wealth management industry.

## Answer by Alex Rodriguez (score -1)

https://quant.stackexchange.com/a/21414

Many use systemic trading strategies that buy/sell according but not limited to:

- P/E - price to earning

- P/B - price to book

- Dividend size

- STD Deviation - volatility

- Sharpe Ratio - return factoring volatility of returns in

- Interest Rates

- Credit Ratings

- Economic indicators (GDP growth, CPI/inflation, Housing data, PMI)

- other various financial ratios for pricing and measuring risk and return of assets