Both grace period loans and community information show promise in bridging the gap between the high returns to investment found in the grant experiments and the lack of a similar effect found in the microcredit experiments. But underlying the lack of appeal to lenders of the grace period loans is the fact that lenders suffer the downside when the investments fail, but do not capture the upside when the investments are successful. The obvious solution to this is some sort of equity contract that allows some risk-sharing (and reward-sharing) by the investor. There are challenges with microequity contracts, however, given the lack of rigorous accounting and auditing standards and the lack of standard exit strategies for investors in these small firms. De Mel et al. (2018) report on a failed microequity experiment in Sri Lanka. Several other microequity experiments are ongoing in Pakistan, Kenya and Indonesia, though as yet there are no results from these studies.
Angel investors providing equity are active in many lower-income countries. However, the networks of such investors are informal and, as far as we are aware, there has been no analysis of their outcomes. Angel and venture investors provide a combination of capital, personalised mentoring and other inputs. Identifying the effects of these contributions on firm growth is particularly challenging because these investors combine careful selection of enterprises with post-selection interventions. Research that credibly separates the selection effect from the investment effect is rare. One interesting attempt to isolate the investment and mentoring effects, albeit using data from the United States, is reported on by Kerr et al. (2011). The researchers use data from internal records of two angel investor groups in southern California and Boston. The two groups use similar methods to select investees, with entrepreneurs making presentations in the presence of many angel investors who are members. The records of Tech Coast Investors, the Californian group, include internal discussions on each of these ‘pitches’ showing the level of support for each pitch. Kerr et al. show that for the majority (64%) of the ideas pitched to the group, no investor is interested. The data also show that the probability of receiving funding jumps significantly after 20 angels express an interest in funding the venture – while 38% of ventures with interest expressed by between 20 and 24 angles receive funding, only 17% of those with interest expressed by between 15 and 19 angels receive funding. Kerr et al. use this discrete jump in funding rates to compare the trajectories of 46 enterprises that were supported by 20-34 angels with rates of funding) with 41 that were supported by 10-19 angels with much lower rates of funding). A similar discontinuity is used to split the 43 pitches to the Boston angel group that were close to a funding threshold. The logic behind this ‘regression discontinuity’ approach is that the pitches falling just above or below the chosen thresholds are likely to be similar in potential but very different in terms of the likelihood of receiving angel support. They will therefore differ in outcomes only (or at least mainly) because the group just above the cut-off received funding and related assistance from the angel group at much higher rates than the group just below.
The analysis by Kerr et al. shows that these two US angel investor groups have an important impact on the trajectory of enterprises. Data from four years after the funding decisions reveal that those receiving funding over 20% more likely to have survived, and over 16% more likely to have either had a successful exit or have grown to at least 75 employees. We are unaware of any similar research in lower-income economies.
