NOTE: The portfolio update went live in December 2013.

At Betterment, we aim to create the optimal investing service that strives for the best-possible results. At the heart of this is — very simply — better expected returns at every level risk.

How much better? Overall expected returns may improve as much as 0.57 percent on average compared to the existing portfolio. In actual dollar terms, that means a $100,000 portfolio invested at the average risk level could see improvement in expected returns up to $51,966 over a 20-year period.¹

Several improvements working in tandem should drive the boost in expected returns: A smarter way of allocating assets at every risk level, a better bond basket, and lower underlying asset costs.

Smarter Allocation

At all risk levels, we are improving expected returns through a more effective use of underlying assets. Customers will still see a stock-to-bond allocation for each goal, but our algorithm will carefully vary the weights of the ETFs to optimize the risk and return for every asset allocation.

The following chart demonstrates the weighted asset allocation. Each vertical slice represents the allocation among the assets for that overall stock-to-bond allocation.¹

Currently, Betterment recommends a higher allocation to the bond basket for shorter term goals, but the bond basket’s composition is the same, whether your portfolio has 10 percent bonds or 90 percent bonds. Flexible allocation goes a step further: the precise composition of the stock and bond basket will now depend on your stock-versus-bond allocation. Lower risk still means more bonds, but for lower risk goals, the lowest risk bond ETFs take on a greater role within the bond basket.

On the other side of the spectrum, for longer-term goals, the higher risk/higher expected return bond ETFs grow as a percentage of the bond basket. The same goes for stocks: at lower risk levels, less volatile Large Cap U.S. stocks make up a relatively larger proportion of the stock basket than they do at high-risk levels.

For example, a goal with a 5 percent stock allocation — the lowest risk portfolio we recommend — will have 65 percent Short-term Treasuries (SHV). A goal with 41 percent or more stock will have no SHV at all (though up to 59 percent of the goal is still bonds). A 90 percent stock goal will have more emerging markets stocks and bonds as a percentage of the stock and bond basket, respectively, than a 60 percent stock goal. What this amounts to is more optimization towards your level of risk: the baskets are composed to better serve their relative roles.

Better Bond Basket

While our new portfolio is designed to outperform the old portfolio at every risk level, it particularly shines at the very low-risk levels because of our better bond basket. Even long-term goals eventually become short-term goals, and this gives us the ability to serve our customers through the entire journey. You can read more about the specific bond ETFs we are adding to our portfolio here.

Lower Cost

Finally, we are also improving your expected returns by keeping underlying costs low. The new blend of funds will have lower underlying costs, by about 0.015 percent on average. In a $100,000 portfolio invested in 50 percent stocks over 20 years, that compounds up to $1,084. Not only is the expected risk-adjusted performance better, but you’ll keep more of those improved expected returns.

Meanwhile, as a customer, you don’t need to do a thing. We are making these changes as part of our mission to keep improving our services.

¹ We employed the Black-Litterman model to generate forward-looking expected returns. To do this, we analyzed the global portfolio of investable assets and their proportions. The Black-Litterman model, a complex mathematical equation, allows us to use this global portfolio to generate forward-looking expected returns – for each asset class.

While many firms also use Black-Litterman to make short-term market-timing decisions by imposing their own views, we adhere to our core index-based investment philosophy and do not. We get the best possible estimates of expected returns – those aggregated from millions of investors across the globe. Please see our full disclosure of our methodology for model returns.