Helping graduated borrowers through asset-based microfinance: Evidence from Pakistan


Published 10.05.21
Photo credit:
Karandaaz Pakistan/flickr

Borrowers who receive loans for a fixed asset run larger businesses and see higher profits, with positive impacts on household consumption

Editors’ note: To know more about asset-based Microfinance, read our VoxDevLit on Microfinance.

Most microfinance loans are relatively small (Banerjee et al. 2015). Their size and structure can be something of a mystery when compared to business financing in many parts of the world, where small firms often have access to contracts that specifically allow them to purchase large fixed assets, often with flexible repayment terms and asset ownership options. This is a stark contrast to most low-income countries, where microenterprise owners typically only have access to inflexible debt-based cash loans that may be insufficient for large fixed capital purchases. Microfinance institutions may be uncomfortable lending larger sums, especially in the absence of appropriate collateral, and if clients do not have productive investment opportunities to generate high returns.

The first wave of microfinance randomised controlled trials (RCTs) found modest average impacts of conventional microcredit contracts on microenterprise performance, and practically zero effects on household consumption (Duflo 2020, Meager 2019). Subsequent work has identified significant heterogeneity in business impacts (Banerjee et al. 2019) and several papers show benefits from contractual innovations designed to increase repayment flexibility (Field et al. 2013, Battaglia et al. 2018, Barboni and Agarwal 2018). In their seminal review of the experimental literature, Banerjee et al. (2015) recommend that the next generation of microfinance research explore contractual innovations and non-credit structures, while addressing the lack of evidence for the impact of larger financing amounts on graduated borrowers.

Experimental design

Following this recommendation, our recent study (Bari, Malik, Meki, and Quinn 2021) explores these ideas through an RCT in Lahore, Pakistan. We worked with Akhuwat—one of Pakistan’s largest microfinance institutions (MFI)—to ‘strongly back’ some clients with loans much larger than the norm. Specifically, we worked with graduated borrowers, or urban and semi-urban entrepreneurs who had successfully completed at least one loan cycle and reached the maximum borrowing amount (about US$ 475 at the time). Instead of offering the borrowers another loan of US$ 475, we offered financing for the purchase of an asset worth up to four times the existing loan limit (US$ 1,900), or just under five times the average monthly household income for this sample.  

To make this feasible, financing was offered using a ‘hire purchase’ structure: by purchasing a fixed asset of the entrepreneur’s choosing and having them repay over the following 18 months. In some countries, this kind of contract is known as ‘rent to own’ or ‘purchasing by instalments’.,

Table 1 shows a simple example of how the repayment contract was structured, imagining a client purchasing an asset worth 100,000 rupees. This requires an initial deposit of 10,000 rupees and financing of 90,000 rupees, after which the client repays 5,000 rupees each month towards the ownership of the asset, and pays rent on the share of the asset owned by the lender (at a monthly rental rate of 1%).

Table 1 Contract structure: Fixed-repayment contract

We also offered a flexible version of the product, in which the client could choose each month how much to repay, with the ownership share varying accordingly. There was a minimum monthly repayment requirement of 2.5% of the total purchase price, or half of that in the fixed-repayment contract.  

Take-up of the asset-based microfinance product

We found high take-up for both variations of the contract. This was high relative to most previous microfinance studies, although our sample of graduated borrowers is not directly comparable to studies that offered microfinance to new clients. Specifically, 53% of those offered the fixed-repayment contract took it up (that is, they provided the required deposit and purchased an asset under the fixed-repayment terms). Of those offered the flexible-repayment contract, a similar proportion—50%—took up the contract. In addition, 9% of those offered the flexible-repayment contract decided to proceed under the fixed-repayment terms. Respondents chose a variety of fixed assets, where the most common included auto-rickshaws, high-quality sewing machines and cameras, and various manufacturing-related machines. We did not observe a significant difference between the value of assets purchased under the flexible-repayment contract than under fixed repayment.

The MFI’s (Akhuwat’s) administrative data reveals that both contracts performed well, with high repayment rates (less than 4% default, with no major difference between the two contracts), and the MFI recovering all of its initial capital. Clients in the flexible-repayment contract generally paid substantially more than the minimum required, and we see significant month-to-month variation in repayments made under the flexible contract, mostly lying in between what entrepreneurs were required to pay and what the equivalent required payment under the fixed contract would have been. This is consistent with Battaglia et al. (2018), who find that the grace periods they offered were used across the loan cycle and sometimes not at all.

Impacts of the product on clients’ enterprises and households

We implemented our experiment with around 750 respondents. Of these, approximately one third were assigned to the control group (eligible for their usual microfinance loan), one third were offered the fixed-repayment asset finance contract, and one third were offered their choice of either the fixed- or flexible-repayment contract. 

To identify the impact of the contracts, we estimate a pooled intent to treat by comparing outcomes for respondents offered a contract to those in the control group. To do this, we pool data from face-to-face surveys conducted three, six, 12, 18, and 24 months after treatment. In our paper, we  disaggregate results over those separate waves and do not find important differences by survey time. Notably, as clients of Akhuwat, the control group were eligible for a zero-interest loan, already much more generous than most products available in the market. In this sense, we deliberately set a high bar against which to draw comparisons.

We find large, significant benefits from our contracts. First, we find a large and sustained increase in total business assets (an increase of about 40% on the control group mean) and on business profits (an increase of about 11% on the control group mean). In turn, this led to increased household income (an increase of 8%) and consumption expenditure (6%). When we disaggregate the increase in consumption, we find this is largely driven by greater spending on food, and by a 26% increase in spending on children’s education, particularly for girls (similar to Jack et al. 2019).

To understand these results, we provide in the paper a calibrated dynamic structural model, in which a household owns a microenterprise and decides on investment and consumption. Specifically, we allow the household to invest either in a low-return liquid asset (for example, cash), or a high-return illiquid asset (for example, an auto-rickshaw).  We model that illiquidity as arising, in part, from capital being ‘lumpy’. Formally, as in Field et al. (2013), we model this by imposing a minimum investment size. This captures the intuition that, for example, an entrepreneur cannot purchase an auto-rickshaw one wheel at a time. In order to enjoy the high returns from fixed capital investments, the entrepreneur must accumulate a large cash lump-sum or fund that lump-sum purchase through a microfinance contract. Our model calibration fits well the observed distribution of data for fixed capital, microenterprise profits, household consumption, and liquid capital; in doing so, it implies an important role for capital lumpiness. 

Implicit insurance through repayment flexibility

Although we find no overall difference in impacts between the fixed and flexible contracts, heterogeneity analysis with pre-specified variables reveals the importance of risk preferences. Using an incentivised measure of baseline risk aversion, we find that the most risk averse individuals had significantly higher take-up of the flexible repayment contract compared to the fixed repayment contract. They are more likely to use the flexible repayment option when faced with business shocks and eventually benefit more from the flexible contract in terms of business and household outcomes (when compared to similarly risk averse individuals only offered the fixed repayment contract). This suggests that the implicit insurance in the flexible-repayment contract is particularly valuable for risk averse microenterprise owners. 


Previous studies that have provided poor individuals in low-income countries with a large capital injection (usually in the form of productive asset grants) have found substantial and persistent increases in business and household income (see, in particular, De Mel et al. 2008, Fafchamps et al. 2014, Banerjee et al., 2015, Hussam et al. 2017, Bandiera et al.2017 and Crépon et al. 2020). While our paper varies somewhat, as our capital injection is provided as a loan recovered by an MFI, our results nonetheless suggest that asset-based financing may provide a sustainable mechanism to generate similarly high returns—at least for graduated borrowers.

Our results show that an asset-based microfinance product can work from both an administrative perspective (high repayment rates, active and intuitive use of the flexible-repayment option), and positive client level outcomes (both business performance and household welfare). Together, our results suggest that large asset-based microfinance contracts may stimulate microenterprise growth among graduated borrowers. Given their Shariah-compliant contractual form, such contracts are likely to have particular appeal for many Muslim clients—a group disproportionately represented among the world’s poor and financially excluded (El-Gamal et al.2014). There is also no reason for these advantages to be limited to Islamic contexts. Asset-based financing is an important source of credit for small businesses around the world, and a form of contract that could readily be extended to many microenterprises.

Editor's Note: This article is based on this PEDL research.


Bandiera O, R Burgess, N Das, S Gulesci, . Rasul and M Sulaiman (2017), “Labor Markets and Poverty in Village Economies”. The Quarterly Journal of Economics 132(2), 811–870.

Banerjee A, D Karlan and J Zinman (2015), “Six Randomized Evaluations of Microcredit: Introduction and Further Steps”. American Economic Journal: Applied Economics 7(1), 1–21.

Banerjee A. E Duflo, N Goldberg, D Karlan, R Osei, W Parienté, J Shapiro, B Thuysbaert and C Udry (2015), “A Multifaceted Program Causes Lasting Progress for the Very Poor: Evidence from Six Countries”. Science 348(6236), 1260799.

Banerjee A, E Breza, E Duflo and C Kinnan (2019), “Can Microfinance Unlock a Poverty Trap for Some Entrepreneurs?” Working paper.

Barboni G and P Agarwal (2018), “Knowing What’s Good for You: Can a Repayment Flexibility Option in Microfinance Contracts Improve Repayment Rates and Business Outcomes?” Working paper.

Bari F, K Malik, M Meki and S Quinn (2021), “Asset-based microfinance for microenterprises: Evidence from Pakistan.” CEPR Discussion Paper No. DP15768.

Battaglia M, S Gulesci and A Madestam (2018), “Repayment Flexibility and Risk Taking: Experimental Evidence from Credit Contracts.” Working paper.

Crépon B, M El Komi and A Osman (2020), “Is It Who You Are or What You Get? Comparing the Impacts of Loans and Grants for Microenterprise Development.” Working paper.

De Mel S, D McKenzie and C Woodruff (2008), “Returns to Capital in Microenterprises: Evidence from a Field Experiment”. The Quarterly Journal of Economics 123(4), 1329–1372.

Duflo E (2020), “Field Experiments and the Practice of Policy”. American Economic Review 110(7), 1952–73.

El-Gamal M, M El-Komi, D Karlan and A Osman (2014). “Bank-insured RoSCA for Microfinance: Experimental Evidence in Poor Egyptian Villages”. Journal of Economic Behavior & Organization 103, S56– S73.

Fafchamps M, D McKenzie, S Quinn and C Woodruff (2014), “Microenterprise Growth and the Flypaper Effect: Evidence from a Randomized Experiment in Ghana”. Journal of Development Economics 106, 211– 226.

Field E, R Pande, J Papp and N Rigol (2013), “Does the Classic Microfinance Model Discourage Entrepreneurship among the Poor? Experimental Evidence from India”. American Economic Review 103(6), 2196–2226.

Hussam R, N Rigol and B Roth (2017), “Targeting High Ability Entrepreneurs Using Community Information: Mechanism Design In The Field”. Working paper.

Meager R (2019), “Understanding the Average Impact of Microcredit Expansions: A Bayesian Hierarchical Analysis of Seven Randomized Experiments”. American Economic Journal: Applied Economics 11(1), 57–91.