Bulletin – June Quarter 2010
Reform of the ATM System – One Year On
Brendan Filipovski and Darren Flood*
The payments industry, with support from the Reserve Bank, introduced major reforms to the ATM system in March 2009. The reforms were designed to increase competition and efficiency in the ATM system. They have, by and large, been meeting their objectives. Cardholders have responded to clearer price signals by changing their pattern of ATM transactions, resulting in a reduction in total ATM fees paid. Flexibility in ATM pricing has helped to increase the number of ATMs, including in locations that might not have been financially viable in the past. At the same time, competition has driven financial institutions of all sizes to provide their cardholders with access to a wide network of ATMs free of direct charges.
In March 2009, a package of reforms designed to improve competition and efficiency in the Australian ATM system came into effect. These reforms changed features of the system that were previously largely hidden from view of the general public and prevented active competition, hindering the industry’s flexibility to respond to changing circumstances. The most visible elements of the reforms were the removal of ATM ’foreign fees’ and their replacement with direct charges. Foreign fees were levied by a cardholder’s financial institution for transactions at ATMs the institution did not own, while direct charges are levied by the ATM owner.
This article reviews the effects of the move to direct charging after a year of operation. While the competitive response to the reforms is still evolving, the Bank’s assessment is that the key objectives of the reforms are being met. In particular, cardholders’ reaction to the increased transparency of fees has resulted in a significant reduction in the fees paid in aggregate on ATM transactions. At the same time, there is evidence that pricing flexibility has resulted in more widespread availability of ATMs, including in rural and regional areas and in locations that are more difficult to service. Competition has driven financial institutions to provide their customers with access to a wide network of ATMs free of direct charges, but at this point there is less evidence of vigorous price competition between ATM owners. This may change as cardholders become more aware of differences in pricing and ATM owners compete more actively to attract transactions.
Direct Charging at ATMs
The key element of the reforms in March 2009 was a change in the way ATM owners are paid for a cash withdrawal or balance enquiry made by a customer of another financial institution (known as a ’foreign’ transaction). Prior to the reforms, the cardholder’s financial institution paid a fee (known as an interchange fee) to the ATM owner. Typically, the cardholder’s financial institution recouped this amount (and often more than that) from the cardholder in the form of a foreign fee. The reforms abolished interchange fees and allowed ATM owners to charge customers directly for the use of an ATM at the time of the transaction. The reforms mandated that the direct charge be displayed to customers prior to them completing the transaction, and that customers be given the opportunity to cancel the transaction without cost if they do not wish to proceed. There were several main anticipated benefits of the move to direct charging:
- Increased transparency of ATM fees: cardholders see the ATM fee at the time of the transaction, rather than when they receive their monthly or quarterly account statement. It was anticipated that this greater transparency would lead to more cardholders taking steps to avoid fees on foreign ATM transactions.
- Greater competition in the setting of ATM fees: ATM owners have the flexibility to set fees in a way that allows them to compete for ATM transactions, while customers are more conscious of the fees they are paying and are therefore more likely to respond to differences in fees.
- Increased ATM deployment: pricing of transactions can better reflect the costs of deployment, making deployment of ATMs viable in a wider range of locations than has previously been the case.
The reforms also included a number of elements that make it easier for potential competitors to access the ATM system. More information on these elements of the reforms can be found in An Access Regime for the ATM System (RBA 2009).
The following sections describe the effects of the ATM reforms on transparency, pricing, customer behaviour, competition and ATM deployment.
Transparency and Pricing
The move to direct charging had the potential to make ATM fees significantly more transparent to cardholders, provided that financial institutions’ foreign fees were removed at the same time. Foreign fees had typically been around $2.00 per transaction prior to the reforms, both for withdrawals and balance enquiries (Table 1). The Reserve Bank made it clear in the lead-up to implementation of the reforms that it saw no case for the retention of these fees since, with the removal of interchange fees, the cost to an institution of its cardholders making foreign transactions would be little different from transactions on its own ATM network. Within a few weeks of the new arrangements, most institutions had removed foreign fees. So, from the cardholder’s perspective, foreign fees typically have been replaced by more transparent direct charges.
|Bank of Queensland||$2.00||$2.00|
|Greater Building Society||$1.50||$1.00|
In most cases, the direct charges applied by ATM owners under the new regime have mirrored the typical level of foreign fees prior to the reforms (Graph 1, Table 2). Approximately 88 per cent of ATMs, including those owned by three of the major banks and most ATMs owned by three of the largest independent deployers, now apply a $2.00 direct charge for a withdrawal. Of the remainder, the largest share charge $1.50 for a withdrawal, including ATMs owned by the National Australia Bank. Around 1,000 ATMs owned by independent ATM deployers (about 4 per cent of all ATMs) apply a $2.50 charge for a withdrawal. A range of other charges apply at small numbers of ATMs. The Bank’s liaison with ATM owners suggests that the highest charges are around $4.00 on a small number of machines in specialised venues. At the other end of the spectrum, in a small number of cases, owners of the venues in which the ATMs are placed are absorbing costs so that consumers pay no direct charge at all. Prior to the reforms, some commentators had suggested some ATM owners might vary ATM charges according to the time of day. This has not occurred to date.
|Cash withdrawal||Balance enquiry|
(a) Also deploys ‘branded’ ATMs for other institutions, which
may apply different charges
(b) Predominantly $2.00
(c) At the discretion of sub-network members, but withdrawals capped at $2.00 and balance enquiries capped at $1.00
(d) $0.50 for withdrawal and $0.25 for a balance enquiry at one remote Northern Territory ATM
|Bank of Queensland||$2.00||$2.00|
Because ATM balance enquiries do not incur the costs associated with cash handling, they are cheaper to provide than ATM withdrawals. Nonetheless, balance enquiries also typically attracted foreign fees of $2.00 prior to the reforms. There is more variation in post-reform direct charges for balance enquiries than is the case for withdrawals, with around three-quarters of ATMs charging $2.00 (Graph 2). Most of the remaining ATMs charge from $0.50 to $1.00 – most commonly $1.00. Around 1 per cent of ATMs do not levy a direct charge for balance enquiries.
In summary, a typical foreign ATM transaction costs the same now as prior to the reforms. Importantly, however, cheaper foreign transactions are possible (particularly for balance enquiries) if cardholders are prepared to seek them out. At only around 4 per cent of ATMs would a typical cardholder pay a higher fee than prior to the reforms and only in a small number of cases is there likely to be no genuine alternative to using that ATM.
Cardholders’ Response to Increased Transparency
While the prices of most foreign ATM withdrawals are similar to those before the reforms, consumers as a whole are paying fewer ATM fees. The increased transparency of the new arrangements, combined with arrangements put in place by individual institutions to provide additional charge-free ATMs to their customers, has resulted in cardholders changing their behaviour in order to reduce the fees that they pay.
In the first year of the new regime, the number of foreign withdrawals fell by 18 per cent from the preceding year, while cardholders’ withdrawals at their own financial institutions’ ATMs (’own’ transactions) increased by 9 per cent. As a share of all ATM transactions, foreign transactions fell from around 44 per cent to 38 per cent and this share has been relatively steady since the reforms became effective (Graph 3). This points to a distinct change in cardholder behaviour.
This reduction in the number of foreign transactions does not in itself, however, provide an indication of the fee savings to cardholders. To estimate this, allowance must be made for cases where foreign transactions do not incur a fee. These cases include accounts on which no foreign fees were levied by the cardholder’s institution prior to the reforms and post-reform agreements that provide cardholders with charge-free access to some ATMs not owned by their financial institution (see below). Allowing for these cases, and similar arrangements between merged financial institutions, the reduction in ATM withdrawal fees paid by cardholders in the first year following the reforms is likely to have been around $120 million. Around two-thirds of all withdrawals are estimated not to attract a direct charge.
Not only have consumers responded to more transparent pricing by using their own institution’s ATMs in place of foreign ATMs, they have also increased the average value of each ATM withdrawal so that they visit foreign ATMs less often. Overall, the number of ATM withdrawals fell by around 3½ per cent in the year after the reforms, even though the total value withdrawn was little changed from the preceding year (Graph 4). This effect has been concentrated in ’foreign’ rather than ’own’ transactions; the average value of a foreign ATM withdrawal was $156 in the year to February 2010, up from $148 in the year prior to the reforms.
The decline in the number of ATM withdrawals also reflects cardholders responding to clearer pricing. Since the reforms were implemented, cardholders have made greater use of EFTPOS cash-outs as an alternative to foreign ATMs, given that these are typically free to the customer.1 The number of EFTPOS cash-out transactions increased sharply in March 2009 and was 9 per cent higher in the first full year of the reforms, compared with the previous year (Graph 5).
Finally, there is evidence that cardholders have reacted strongly to direct charges on ATM balance enquiries. Information provided by financial institutions and ATM owners suggests that balance enquiries fell by one-third to one-half when the reforms were introduced. This strong reaction suggests that some cardholders were not aware of the foreign fee for these transactions and have sought to limit their use now fees are transparent.
Effects on Competition
The reforms have affected competition between different types of participants in the ATM system in different ways. For financial institutions, the priority has been ensuring that their customers have ready access to their funds free of charge via a wide network of ATMs. For independent ATM owners, on the other hand, the focus has been on generating profits from their ATM network. The pricing flexibility available to ATM owners following the reforms provides scope for them to compete for business more actively.
Competition among financial institutions
Access to deposits through a network of ATMs is an important part of the service that most financial institutions provide to their customers. Financial institutions have generally continued to provide their cardholders with free withdrawals at their own ATMs, as they did prior to the reforms, while direct charges apply to other transactions.
One concern about a direct charging model, however, is that it could potentially put smaller institutions at a disadvantage because they typically have fewer ATMs than the large banks so that their customers have fewer opportunities to make ATM withdrawals free from direct charges. Given this, and in order to promote competition, the reforms allowed for financial institutions to pay interchange or similar fees to ATM owners in limited circumstances in lieu of the owner applying a direct charge. In this way, institutions could provide their customers with access to additional ’charge-free’ ATMs.
One option available to smaller institutions is to be part of an ATM sub-network, where a number of institutions effectively pool their ATMs under a common brand and allow one another’s cardholders charge-free access to the network’s ATMs (examples are the RediATM and FeeSmart networks). Another option is for a financial institution to enter a ’one-way’ agreement with an individual ATM network owner to provide its customers with access to those ATMs either charge-free or at a reduced charge. A third option is for an institution to outsource its ATM fleet by paying the operator of an independent network to place the institution’s branding on some ATMs. These ATMs would continue to be owned and operated by the independent deployer, but would have the outward appearance of being owned by the financial institution, with the institution’s cardholders paying no direct charge.
Many institutions have made use of these options so that their customers have access to a large network of charge-free ATMs. Of the largest 30 financial institutions in terms of ATM usage, 28 institutions (covering 98 per cent of all ATM cardholders) are providing access to a network of at least 1,500 ATMs charge free or at reduced charges to their cardholders. The remainder tend to have a geographically concentrated base of account holders and are able to offer charge-free transactions in the local area.
In many cases, sub-networks, one-way agreements and outsourcing mean that cardholders have access to more fee-free ATMs than prior to the reforms, but there are also some cases where access to free ATMs has been reduced. Some institutions – in particular some credit unions and banks that do not have large branch networks – previously did not levy any foreign ATM fees, choosing instead to absorb all ATM costs for their customers. Under the new regime, most of those institutions have decided to meet their customers’ costs only for a single foreign ATM network.
While the number of financial institutions offering unlimited free access to ATMs has been reduced, the reforms do not prevent an institution from continuing to do so by absorbing ATM direct charges on behalf of customers. At least two banks – Bankwest and ING – offer accounts that provide charge-free access, either to all ATMs or to those owned by the major banks, by rebating any direct charges incurred by their customers.
Competition among ATM owners
Under the reforms, ATM owners are freely able to set their price in a similar way to most other firms in the economy. As discussed above, there has been some variation in the direct charges applied, but the $2.00 foreign fee that applied prior to the reforms has established a benchmark. While few ATM owners have moved very far from this benchmark to date, over time some ATM owners might seek to generate additional revenue by adjusting the level of direct charges. One possibility is that owners of ATMs that are in close proximity to others will try to attract traffic by lowering charges.
While cardholders are making greater use of their own institutions’ ATMs or EFTPOS cash-outs to avoid ATM charges, where they use a foreign ATM, there is little evidence that they seek out the ATM with the lowest charge. In other words, where an ATM charging $2.00 and another charging $1.50 are in close proximity, cardholders do not necessarily choose the ATM charging $1.50. It is possible that this reflects the fact that cardholders typically need to proceed some way through the transaction process at the ATM before the direct charge is displayed. This may make it difficult to compare prices, particularly where cardholders are unfamiliar with the ATMs in a particular location.
While cardholders might avoid an ATM if they see that it applies a direct charge higher than they think is reasonable, there is little incentive for ATM owners to lower fees if it is not obvious to potential customers that they have done so. As a consequence, relatively little price competition among ATM owners appears to have developed to date. An obvious response is for ATM owners with a low direct charge to advertise the charge prominently so as to attract additional throughput and higher fee revenue. There is nothing to prevent owners from doing this, although the strategy requires the general public to understand pricing sufficiently to react accordingly. To date, advertising of prices has occurred only in isolated cases, but this and other competitive strategies might develop over time as the market matures. The presence of different types of ATM owners might assist this process. Overseas, some financial institutions reduce direct charges on their ATMs so that they can attract their competitors’ customers and expose them to their own advertising through the ATM.
Deployment of ATMs
The flexibility that ATM owners now have in setting ATM prices means that the payments they receive for use of their machines can better reflect the costs of deployment. These costs have been rising over time. A study by the ACCC and the Reserve Bank in 2000 estimated that the weighted-average cost of an ATM transaction at that time was 49 cents (RBA and ACCC 2000). A 2007 Reserve Bank study estimated that the weighted-average cost was 74 cents and the median 85 cents (Schwartz et al 2008). While the methodology of the two studies differs somewhat, the 2007 study concluded that costs had clearly risen for some inputs, including cash-handling and site rental. One of the objectives of the reforms was to provide greater flexibility in ATM pricing so that ATMs could remain financially viable, particularly in more costly locations and circumstances. Without that flexibility, there was a risk that the number of ATMs would begin to decline over time if the cost of operating existing machines rose too far relative to the interchange fee. Machines in more costly locations would be most at risk, including those in rural and regional areas and those requiring greater security.
The year-to-year volatility in provision of machines makes it difficult to draw strong conclusions about post-reform ATM deployment relative to earlier periods. Nonetheless, the evidence suggests that the reforms have been positive for the availability of ATMs. Data reported to the Reserve Bank suggest that ATM numbers have increased by about 1,500, or about 6 per cent, under the new regime. Of these, about half have been deployed by financial institutions and half by independent deployers, broadly in line with ownership of the network overall (Table 3). Moreover, the Bank’s liaison suggests that some of these ATMs have been deployed in locations that might otherwise have not been viable – including in rural, regional and remote areas. It is also becoming more common to see ATMs in relatively low-usage locations and temporary ATMs at public events. Such ATMs tend to apply above-average direct charges, but would most likely not have been available under the previous regime where owners typically received $1.00 per transaction by way of the interchange fee. That said, deployment has been uneven among networks, with at least one provider reporting a reduction in ATM numbers in response to lower transaction volumes.
|(a) Some figures include ATMs carrying financial institution
branding, but owned by an independent deployer. These may be recorded against both the owner
and the branding institution.
|Commonwealth Bank & Bankwest||3,714|
|Bank of Queensland||3,577|
|RediATM (including NAB)||3,171|
|Westpac & St. George||2,971|
A number of new independent deployers have also entered the market. These players are helping to generate vigorous competition for ATM locations, with one consequence reportedly being rises in rents for ATM sites. The models used by independent deployers vary. In some cases, sites are rented by the deployer, which retains the direct charge, while in others it has become common for the direct charge to be shared in some way with the site owner. The machines themselves may be owned either by the deployer or the site owner. Another recent model has been for the deployer to sell ATMs at third-party locations to investors, taking advantage of the government investment tax incentives that were in place until the end of 2009.
One issue raised by groups representing visually impaired people prior to the implementation of the reforms was a concern about the availability of audio-enabled ATMs and the potential impact of direct charges on people who are limited in the ATMs they can use. The Bank’s liaison with the industry indicates that around 30 per cent of ATMs are now audio-enabled.
The changes introduced to the ATM system in March 2009 are, by and large, meeting their objectives. ATM fees are much more transparent and cardholders have responded to clearer price signals by changing transaction patterns in a number of ways that have allowed them to avoid or reduce ATM fees. In aggregate these changes have resulted in a reduction in ATM withdrawal fees paid of around $120 million in the first year of the new arrangements. This has occurred even though the typical post-reform direct charge is unchanged from the level of the typical pre-reform foreign fee, with significant price competition yet to develop among ATM owners. There has also been an increase in the supply of ATMs and ATMs are becoming available in locations and circumstances where they would not have been viable under the previous arrangements. Competition among financial institutions has worked to provide the vast majority of cardholders with access to a wide network of ATMs free of direct charges.
* The authors are from Payments Policy Department. [back to text]
- The EFTPOS system allows cash to be withdrawn from a deposit account at an EFTPOS point-of-sale terminal. This can occur either in conjunction with a purchase or as a separate transaction.
RBA (Reserve Bank of Australia) (2009), An Access Regime for the ATM System, Sydney, February.
RBA and ACCC (Australian Competition and Consumer Commission) (2000), Debit and Credit Card Schemes in Australia: A Study of Interchange Fees and Access, Sydney, October.
Schwartz C, J Fabo, O Bailey and L Carter (2008), ‘Payment Costs in Australia’, Payments System Review Conference, Reserve Bank of Australia and Melbourne Business School, Sydney, April, pp 88–138.