We have a strong incentive and desire to transfer the majority of the funds we receive to our recipients as quickly as possible. GiveDirectly is a service – donors use GiveDirectly to transfer funds to the extreme poor – and our donors expect that service to be efficient and timely. And beyond that, transfer speed is important because according to research cash transfers also offer our recipients a significantly higher rate of return than we could hope to get ourselves by saving or investing, and then sending these funds at a later date.

Every day that a donation sits in our bank account rather than in the hands of our recipients the world is effectively losing the difference between the rate of return our recipients would receive, or about 30%, and the rate our investments are paying us. In many cases, that means transferring money quickly and directly is the best possible option.

Image not found
Rispa used her GiveDirectly cash transfer to invest in livestock, maize, and beans for her farm. A randomized controlled trial on GiveDirectly’s recipients showed that income from businesses and agriculture increased by 28% of the average grant size, implying a 28% annual rate of return.

Therefore, when a donor gives us money we make our best efforts to get that into the hands of the poor as quickly and efficiently as possible. However, sometimes we receive grants from larger institutions who specify that they are interested in funding our growth in a different way. For example, in 2015 we received a $25 million grant from Good Ventures that was directed at building partnerships with other grant-making institutions and building a world-class fundraising team.

By using those funds as matching funds to leverage partnerships with the largest institutional donors, these partnerships can both multiply the amount of cash transfers going to the poor, and help to reshape the $130 billion aid sector so that it more effectively serves beneficiaries. These negotiations can take longer than our typical collection and distribution process, but if bringing $10 million of our own capital to the table attracts an additional $10 million from another institution, and creates the basis for systemic change, we think it will often be worth the time and effort required.

We also need to carefully manage the rate at which we push out funds, both to check against risks of potential fraud or mismanagement, and to prevent disruptions in our operations due to short-term revenue shocks. In any recession donations can drop significantly, and we hold a small reserve fund to prevent GiveDirectly from having to completely shut down. This amount should always remain as small as possible, and be limited to an amount sufficient to pay the highest priority expenses (e.g. rent, contracts) for a 12-18 month period.

With funds held for these two reasons (partnerships and reserves) we need to decide: What should our risk tolerance be? How should we express that risk tolerance in our operations?

Capital held anywhere has risks. If we left the funds in a savings account they would be losing value by an amount equal to inflation, and if we put all of the money in the S&P500 we would be totally subject to the swings of the US stock market. The likelihood of a positive return in any portfolio relates directly to the expected return, volatility, and holding period of the investment. We decided to build a simple diversified portfolio of assets with a modest expected return, but a relatively high likelihood of positive return over the time period we expected to hold the funds.

Image not found

We started by only investing funds that we expect to hold for at least two years. We then constructed a portfolio of stocks, nominal bonds, inflation-linked bonds, and commodities. Given our relatively low tolerance for risk, we prioritized diversification over absolute total return. To weight the assets in the portfolio we followed something known as the risk-parity approach to portfolio construction, weighting securities based on their risk contribution rather than the capital invested. Opting for simplicity as well, we were able to do so using four mutual funds and one ETF, with total management fees of only 0.16%.

While returns will vary, over the long term we believe this portfolio will return a modest 4% per year. However, it will do so with significantly less volatility than a typical portfolio composed of mostly stocks and nominal bonds. While a traditional portfolio composed of 60% stocks and 40% bonds is expected to have a down year about once every 30 months, we believe our portfolio will only experience an annual loss every 42 months.

While no portfolio is guaranteed to provide a positive return over a specific time period, we believe taking some amount of calculated risk through time will provide GiveDirectly with a positive return sufficient to both offset inflation and increase our operational impact.


Gavin Walsh is GiveDirectly’s Director of Finance and Information Systems. Prior to joining GiveDirectly he spent seven years at Bridgewater Associates in the areas of Trading and Portfolio Construction. He holds a BA in Economics and Psychology from Purchase College, an MS in Quantitative Finance from Fordham University, and the Chartered Financial Analyst designation.

Back to List