“All money is a matter of belief.” This timeless, elegant quote from Adam Smith remains undisputable. The value of any object—whether it’s tangible or intangible, a currency or an asset, or real or imaginary—rests on the collective beliefs of society of what the object’s value is. This fundamental principle holds true for the U.S. dollar, gold, and real estate. It also holds true for digital assets, which have captured the headlines and spawned a whole new industry and asset class.

While everyone may have an opinion about this new asset class, not everyone understands it. And even though there’s a wealth of literature on digital assets for readers with various levels of technical literacy, there hasn’t been any objective evaluation for investors from a wealth management perspective. So should investors include digital assets in their portfolios? And, if so, how should they determine a reasonable allocation? Dr. Randolph Cohen, a senior lecturer at Harvard Business School and a partner at Alignvest Investment Management, addressed these questions at a CFA Society Toronto luncheon seminar last fall.

Cohen began by laying the foundations required for performing an objective analysis first, addressing the complexity and the value drivers that perplex most investors. While digital assets may seem completely different at the surface when compared to traditional assets (e.g., currencies, stocks, or bonds), their core value driver is very similar. All of the assets are mere abstract representations of beliefs that society assigns to them, allowing them to be used for transactions and enabling markets to function. In other words, Cohen explained, the value of digital assets is an abstract representation within distribution ledgers across global networks, while the value of traditional assets is represented by a never-ending chain of belief.

He further clarified that he considers digital assets to be an array of various instruments rather than just Bitcoin, which seems to dominate everyone’s imagination. (There’s actually a much broader universe of digital assets investors should look at.)

However, before considering if digital assets should be included in an investment portfolio, Cohen considered two questions to make the case for their addition. First, is there a risk premium that offers a positive return? And, second, do these assets help grow and protect wealth?

First, in terms of a positive return, Cohen explained that digital assets have some properties that are similar to “liquid venture capital” and “digital gold.” The first term refers to the point of view that digital assets can be thought of as equity or venture capital–like investments in early-stage blockchain technologies in terms of their risk-return profile. Given the relatively short life span of digital assets, there’s not a lot of empirical evidence to prove that, so this argument can be considered as a conceptual construct. However, digital assets do have a point-in-time valuation in that they provide access to the underlying blockchain network, which becomes more valuable as the membership of the network grows.

“Liquid gold” is the concept that digital assets can’t be inflated away like fiat currencies, making them a good store of value. Constraining supply by fixing an absolute maximum quantity or setting a predefined growth rate also supports this argument. Digital assets have structural advantages as stores of value since they’re easier and less expensive to access and store, they’re portable and usable across geographic boundaries, and they can have multiple uses. Again, the more digital assets are used, the stronger the support for their valuation.

As for evidence that digital assets can help grow and protect wealth, these assets are uncorrelated, and, hence, diversifying, relative to all other traditional and alternative assets. Digital assets’ anticipated demand cuts across sovereign boundaries and restrictions, and there are no obvious structural similarities between the returns of digital assets and other asset classes. All of these factors justify including digital assets in a portfolio as a distinct asset class.

So if digital assets should be included in a portfolio, how should investors allocate them? Cohen used the analogy of investing in Internet companies before the dotcom bubble burst. Given the complex and dynamic nature of digital assets today, picking the future winner in this space would be similar to picking Google or Amazon from hundreds of names, each with an unproven value proposition, he said. Assuming that picking individual winners and timing the market can’t be done with any certainty, the best strategy is to passively hold a basket of digital assets with an allocation corresponding to their current market share.

But all of these positive factors don’t negate the evidence that digital assets are volatile and attract a lot of speculative activity. They also have a massive right tail, so investors don’t need much exposure in their portfolios to have an impact. Cohen suggested, based on all of these factors, that it’s completely reasonable to have an allocation of zero percent, and it’s also reasonable to have an allocation of two percent. Allocations of five percent or more would be hard to justify, and would not be prudent. In scenarios where one or more assets performs well, there’s substantial upside to the overall portfolio.

On the flip side, if the value of these assets decreases substantially, the overall impact to the portfolio is immaterial. This skewed set of outcomes justifies a small allocation to digital assets, enough to add a new asset class and uncorrelated risk exposure to investors’ portfolios.

Cohen clearly opened the doors for interested investors to further investigate and focus on the real drivers of value of various digital assets. He hopes, as this asset class matures and its underlying networks become more entrenched, investors will have more empirical evidence to justify the case for including these assets in their portfolios.

This article is based on “Illuminating the Path Forward: Digital Assets in Institutional Portfolios” by R. Stevens, J. Zwick, and R.B. Cohen.