At 8 ATMs per 100,000 people, why is Nigeria not growing her ATM population?

Recently I was reading a World Bank Study on banking penetration and growth. You can see it here. It was fascinating to look at some of the statistics and the conclusions. Let us start first from the statistics — Brazil with a population of 200m has 20 ATMs per 100000 people. See the rest of the chart below.

Put another way, while Nigeria has a total of 15000 ATMs to serve 180m people, Bazil has more than 200,000 ATMs. Compared on a per square kilometer basis, the numbers are even more startling. Nigeria is some where around 0.016 ATMs per square kilometer. If we are to use travel time and cost to estimate, transport costs to the nearest ATM in Nigeria is some where around N1000 on the average i.e. the cost to travel the 10km on average to the nearest ATM ( I am using Rotary club transport policy I found here Copy of Rotary 2016–17 Mileage Reimbursement Rates 1 September …).

The numbers are the numbers. What is more interesting is what the numbers seem to correlate with. One of the conclusions of the study is that ATM penetration seems to go hand in hand with business growth.

“Greater demographic branch and ATM penetration is associated with faster growth rate of firms, as measured by the percentage change in sales over the last three years. Also, the growth rate of firms is higher in countries with higher number of loans per capita. The economic effects of outreach on the growth rate of firms are considerable. A one percent increase in the ratio of branches per capita leads to a 0.17 percent increase, on average, in the growth of sales. A similar in- crease in the number of ATMs per capita results in a 0.27 percent increase in sales growth. Loans per capita appear to have the largest economic impact, with a percent point change inducing almost an average of 0.5 percent change in sales growth.”

Of course, the correlation is not difficult conceptually — the faster and easier it is to reach cash, the faster and easier it is to spend it especially in the very large informal market that dominates Nigeria’s economy. Put another way, deploying more ATMs will led to economic growth. Of course we have to be wary of these correlations because there may be no causal relationship at all. However, we have been for years using similar rationale to push the broad band penetration as there is a correlation between broad band penetration and GDP growth. We can take it as a good reason therefore to pursue ATM penetration. ( One thing with these statistics is that we can come up with all kinds of correlations too).

Looking at the statistics from NIBSS it is interesting that the volume of transactions happening on ATMs have been growing at a significant rate. Much more than any other channel. Well over 80% of the e-payment transactions in Nigeria, happen on ATMs. However, if we look at the population of ATMs deployed in Nigeria, the population has been stagnant over the past 5 years or so. The question is not therefore whether Nigeria needs ATMs or whether having more ATMs is good for the Nigerian economy. The question is why are there no more ATMs in Nigeria than we currently have?

An easy answer of course is that ATMs are expensive to procure and expensive to maintain, and therefore Nigerian banks being shrewd business men have no need to build up cost centers. And this answer is of course correct. At costs of $10,000 per ATM, bridging the gap with a Brazil will cost somewhere in the region of $2bn and a further $200m or more per annum to keep them running. This has not included the additional operational costs around power, communications, cash, security, real estate and what else have you. However, these costs are not that different in other parts of the world including Brazil where one bank alone can have 50,000 ATMs or more. May be the banks in other places are a lot richer than our banks or less shrewd than our own. Richer, most definitely — the third largest bank in Brazil is ten times the largest bank in Nigeria. Less shrewd, I doubt.

In several places in US, Europe and even in India, ATMs and PoS devices are not even deployed by banks alone. It is the mixture of deployers that has led to increased ATM penetration. Independent deployers and ATM networks treat ATM deployment and operation as a business proposition — they provide a value, people appreciate the value and pay for it. And in every business, there is growth when the business is inherently profitable, what users pay exceeds the cost to provide the service. An analysis of the share of market of independent ATM deployers across the globe show clearly that where there is a strong independent deployer presence, the ATM penetration is also high. And independent deployers only get in the game where it is profitable.

It is therefore a plausible hypothesis that the growth in ATM deployments in Nigeria is stunted because it is not profitable to deploy ATMs by either banks or independent deployers in Nigeria. While a bank can use the ATM as a cost reduction and outreach or branding investment, the independent deployer cannot. Even for a bank, there is a limit to how much can be invested in branding. The last time independent ATM deployers were licensed to get into the game in Nigeria, the three licensed IADs ended belly up barely two years after getting their licenses. If we look at the numbers for Euronet, one of the largest ATM networks independent deployers, here , it is obvious that independent ATM deployers can be very profitable businesses except in Nigeria.

ATM deployment in Nigeria is a loss making business because no one pays for the value provided. Short and simple. To see where this is coming from, take a look here . And then consider that in Nigeria, for a bank, ATM transactions are entirely free except if after 3 withdrawals on another ATM other than your bank’s ATM. And even then the fee is a flat N65 — something between 8cents and 20cents depending on the exchange rate we use. Assuming you were charging for every transaction, it will take at least 100,000 transactions to recover the cost of the ATM alone — and not the cost of the communications, power, cash or cash handling. In the US you can pay anything from $3 to $6 per transaction. You can recover your cost of ATM in 2000 transactions which can happen in a few months in a good location.

The fee in Nigeria is of course not coming as a result of the forces of demand and supply or a proper pricing of the value of the service being rendered. It is entirely by fiat, a directive or policy statement issued by the Central Bank of Nigeria — here. This is of course very curious when you realize that the same CBN allows agents to offer cash-out services at a much higher fee of up to N100 per transaction . Of course an ATM is simply a mechanical cash-out agent at it’s simplest configuration. Interestingly of course, the CBN also “approves” that independent ATM deployers can charge up to N150, that is somewhere between 15 and 20 cents, per transaction. Compare this with the N100 that you charge for electronic transactions which are riding on less expensive infrastructure and you begin to understand why payment gateways are appearing from very wood work but no one is investing in ATMs. Compare it with $3 per transaction in US and you dont have to wonder.

The reason why the ATM penetration is not improving in Nigeria is very clear — it is not a profitable business. And it is not a profitable business because the fees paid per transaction is regulated by the CBN and set at an unrealistic level. The entire price setting regime adopted by the CBN like all price setting schemes elsewhere are counter productive and inhibits investments. Market forces will and do moderate forces as has happened in the GSM market. To get the ATM penetration growing and reap the attendant growth in business and GDP, the CBN has to let go and allow market forces attract independent deployers to invest.