The real reason behind high interest rates and what to do about it

The prevailing narrative in the policy discourse on interest rate is that high Non-Performing Loans (NPL) occasioned by macroeconomic volatility, is the primary driver of high interest rates charged by banks on various credit facilities. Factual as that may be, it is becoming increasingly apparent, that operational inefficiency may be a major contributory factor as well. A study of 7 banks using calculated ratios (see Figure 1) from their audited financial statements suggests that high operating cost and weak non-interest revenue are key influencers in determining interest rates on credit facilities.

Banks rely, broadly, on four sources of income to fund operations; interest income, fees and commissions, trading and investments. The debate on interest rate spreads is therefore incomplete without a critical analysis of industry cost behavior. A review of average cost/income ratios in the banking industry shows an upward trend over the last 3 years reaching 52 per cent in 2016 (See PwC Banking Sector Report 2017, p.55). Though Ghana compares favorably against Kenya (62.9%) in terms of industry cost/income ratio, the fact that both economies are currently wrestling with the question of high interest rates suggests something more than sheer coincidence.

Consequently, a critical analysis of the operating cost structure in the banking sector, would in the least shed light on what strategic options are available to management teams as banks pursue earnings growth.

 

Research Approach

To start off, a research hypothesis was needed to anchor a methodical inquiry. The hypothesis is that operating expenses (which inversely correlates with Overhead Absorption Rate), has a greater, if not equal impact (weight) on interest rate charged by a bank on credit facilities.

To arrive at logical conclusions, four things needed to be done:

1. A research of existing literature to understand models for loan asset pricing.
2. Develop a theory as framework for analysis.
3. Collect financial data from a sample of banks for at least 2 years.
4. Analyze the data using the theoretical framework to investigate links, if any, with Annual Percentage Rates charged by banks contained in the sample.

The analytical framework that was used emerged after reviewing some relevant literature, most of which suggested that loan asset pricing accounted for 4 key variables; Loan loss provisioning, target margins, operating cost and funding cost. Thus L_(p) = P_(bd) + R_(sf) + Op _(exp) + F _(mm)

_Where L_{(p) =} Interest rate on facility

P_(bd) = Provision for bad and doubtful debt

R_(sf) = Target rate of return on shareholder funds

Op _(exp) = Overhead Absorption Rate

F _(mm) = Weighted Average Cost of Funding

 

To understand the relative impact of operating expenses [Op _(exp)], as compared to the other price inputs (provisioning, margins and funding cost) on loan asset pricing, the former was deconstructed and expressed as ratio of Operating Income (in most cases) in order to estimate its impact on operational resilience. For instance, General and Administrative Expenses, as a ratio of Operating Income may point to high operating leverage which may in turn point to how a bank’s business model is structured. Again, high rent charges (lease amortization) may, to an extent, indicate high volatility in the property market, but may as well suggest unnecessary profligacy in an era where the market is shifting towards e-channels to deliver consumer value. Other measures of efficiency such as Employee Productivity Index (which affects overheads absorption rate) was also considered, where headcount numbers were available. To compare inter-group performance, simple averages of calculated ratios were used (see Table 1). In all a total 11 metrics (see figure 1) were analyzed using two year data (2015 to 2016) from the audited financial statements of the banks contained in the sample.

 

To help choose a sample, Bank of Ghana’s notices to the public on rates charged by banks, was used as sampling frame. Specifically, NOTICE NO. BG/GOV/SEC/2018/03 was used to select the top four banks with Highest Base Rates (HBR), as a treatment group, and top 4 with Lowest Base Rates as a comparator group. Barclays Bank, with a base rate of 18.1%, was excluded from the comparator group due to unavailability of detailed financial report for period under review. The final sample was therefore reduced to 7. See table 1 for sample details.

Table 1. Sample of Banks included in the study.

Banks With Highest Base Rates (HBR Group)
Name	Base Rate	APR
Premium Bank	34.3%	35.7%
UniBank	33.7%	36.7%
The Royal Bank	32.9%	37.4%
Sovereign Bank	32.2%	33.7%
Banks With Lowest Base Rates (LBR Group)
Name	Base Rate	APR
Standard Chartered Bank	18.2%	28%
Stanbic Bank	18.4%	24.7%
SG Bank	18.9%	29.5%
Notes: 1.      Annual Percentage Rates represents are actual rates charged by banks on loans. Calculation was done using average rate on loans to the Enterprise sector; agriculture, manufacturing, commerce and construction. 2.      BoG Notice BG/GOV/SEC/2018/03 reflects rates as at January 2018.

 

Key Findings and Interpretation

The following key findings were made:

1. By and large, both HBR and LBR groups spent about GH¢0.26 as personnel expense for every cedi earned as operating income. This is GH¢0.10 higher than amount expensed through impairment loss (GH¢0.16) on financial assets.
2. Per the sample reviewed, wages and salaries constitute about at least 70% of personnel cost, with a third allocated for provident fund, medicals and Social Security contributions. The only exception is Sovereign Bank and Premium Bank, who in 2016 allocated 55.2% (GH¢4,517,000) and 43% (GH¢1,765,146) respectively, of personnel budget to wages and salaries. As high as 35.1% was used by Sovereign Bank to fund other staff benefits compared to 29.2% allocated by Premium bank for same purpose.
3. Employee productivity (measured as ratio of operating income per employee to personnel cost per employee) for Unibank (6.2 times) and Premium Bank (5.9 times) were the highest across board in 2016. Interpretation: It simply means that the operating income generated by each of UniBank’s 803 employees in 2016 (GH¢489,321 per employee), was about 6 times the amount of personnel cost per employee (GH¢78,502). In terms of year-on-year employee productivity growth, Premium Bank led with 55.3%. Although the HBR group out-performed the LBR group in this metric, it’s not clear whether the difference is borne of an inherent competitive distinction. In other words, the employees of Unibank and Premium did not perform better on productivity because they were more competent or more motivated. Any meaningful comparison has to account for other contextual factors that contribute to productive capacity, like size of operating asset.
4. On the average, director’s emoluments as a percentage of personnel cost was higher (7.9%) among banks with High Base Rates compared to the portfolio banks in the comparator sample (2016: 4.2%; 2015: 3.9%).
5. The loan impairment problem seem to be a common thread across the industry where highs and lows alternate year-on-year. The impact on operating income nonetheless, differs across the sector. For instance, Stanchart’s GH¢81,687,000 impairment loss in 2016 constituted 13.1% of operating income, down from 40.1% in 2015. On the contrary, The Royal Bank experienced tremendous pressure on income with GH¢61,224,395 impairment loss, which constituted 58.8% of operating income for same period.

 

Trends and Insights

Key differences in terms of cost behavior and revenue structure were found in the sample.

1. The key difference between Highest Base Rate and Lowest Base Rate group is the former’s reliance on net interest income to fund business operations. In 2016, UniBank’s interest expense – GH¢ 1,044,754,294 was 2.7 times its operating income. Similarly, Royal Bank’s 2016 interest expense (GH¢218,733,937) was 2.1 times its operating income. In comparison, Stanchart in 2016 spent GH¢89,108,000 on interest payment, which was 14.5% of operating income. Stanbic Bank spent GH¢109,854,000 on interest payment, representing 18.5% of operating income. The insight is this: comparatively, the LBR group (Stanchart, Stanbic and SG Bank) have more depth in terms of non-interest revenue sources. For instance, Standard Chartered Bank’s net trading income in 2016 was GH¢92.9 million, sufficient to cover an interest expense of GH¢89.6 million. Similarly, Stanbic Bank’s 2016 net fee and commission income (GH¢108.6 million) was just 1.1% shy of total interest expense. This flexibility allows Standard Chartered, Stanbic and other banks with similar revenue structure to set lower base rate on credit facilities, even though customers don’t benefit from this cost-savings because of high risk premiums.
2. Industry expenditures on software is quite high, although like other things, the cashflow impact depends of depth of pocket. In 2016, Sovereign Bank and Premium Bank spent GH¢ 6.3 million and GH¢4.7 million respectively on software. A quick comparison here would give better context. Stanbic Bank within the same period made additions to its intangible assets (software) valued at GH¢9,925,000 representing 1.5% of operating income. Premium Bank and Sovereign bank’s investment in software was 19.5% and 23.8% respectively of operating income. Now, besides the year 1 cashflow impact, amortizations for subsequent years would put more pressure on Premium Bank or Sovereign Bank’s income statements than the case would be for Stanbic. Reducing this adverse impact through higher margins on loans is an option more appealing to the HBR group than the LBR group, for obvious reasons.
3. Per the sample, the LBR group showed better cost efficiency (see Fig. 3). Average General and Administrative Expenses for Stanbic, Stanchart and SG Bank was 12.5% of operating income for 2016 compared to 37.3% for Unibank, Royal Bank, Sovereign Bank and Premium Bank (see Table 1).

 

So what does it all mean?

Although interesting trends have emerged in this study, it may be hasty to suggest that using 11 ratio categories to analyze 7 banks using financial data only, is sufficient to generalize the conclusions across the entire industry. Another reason to be cautious in interpretation is that, the Ghana Reference Rate (GRR) model of loan pricing is different, at least in theory, from the previous model that considered base rates. This study, having relied on January 2018 Annual Percentage Rate (APR) data, has more association with the base rate model of pricing than the GRR. The actual impact on Ghana Reference Rate on banks’ Annual Percentage Rates charged on loans will become clearer as more recent data becomes available.

That notwithstanding, the insights gleaned from this study may be useful in building operating resilience through revenue diversification and rigorous operational cost control. The conclusion made by this study are as follows:

1. Notwithstanding macroeconomic volatility which adversely impacts portfolio quality, strong operating performance depends on having a widely diverse revenue stream.
2. Personnel cost centre is most vulnerable when it comes to cost slippages, particularly discretionary costs such as travel, bonus and perks.
3. High capital expenditures on intangible assets with shorter useful life (3-5 years) adversely impacts revenue performance through high amortizations.
4. The loan pricing behavior of HBR group are more sensitive to high operating costs than LBR group due to the latter’s capacity to generate more non-interest revenue.

 

Implications for corporate strategy

1. Management teams must to track key costs and continually evaluate the impact of new technology investments on operating costs. User data from banking apps must be cross-referenced with cost data associated off-line transactions.
2. The need to also track employee productivity per capita using rigorous metrics, cannot be over-emphasized.
3. The shareholders, through the board of directors needs to hold management accountable for the increasing phenomenon of corporate profligacy. There are efficiency questions to be asked of any management team that would spend 35.1% (GH¢2,873,000) of entire personnel budget on staff benefits, when wages and salaries alone constitutes 55.2% (GH¢4,517,000) of personnel cost.
4. By close of year 2016, SG Bank through a voluntary redundancy program, had reduced headcounts from 914 employees to 728 (25.5%). The impact was a 53.4% growth in operating income per employee (2015: GH¢296,129.31; 2016: GH¢451,167.68). Intelligent reorganization is the new normal.
5. Focus on developing new streams of non-interest revenue from untapped and underserved markets.
6. There is scholarly evidence that employee training and development positively impacts corporate performance in the medium to long-term. Banks needs to shift more resources from perks and benefits to training.

 

Implications for policy and regulations

1. Short of interfering in the strategy formulation process of banks, the regulator must explore creative ways of incentivizing cost control. Although policy tools that rely on moral suasion are less effective, some positive outcomes will be realized in the long term.
2. The growth and development of Ghana’s Fintech ecosystem is directly linked to long term cost curves in related industries. A stronger and competitive local ICT sector will help reduce costs for banks.
3. Addressing volatility on the supply side of property market (both retail and commercial) is of strategic economic importance.

 

To address the high interest rate problem, all the fundamentals that contributes to cost build-up need to be fine-tuned through a coordination policy, regulation and sound management practice. Any regulatory approach that is not sensitive to the dynamics in these three markets; labor, real estate and ICT, would not yield the expected outcomes. Kenya is a prime example, where interest rates caps have resulted in a tightening of credit to the private sector.

About Author

Nkunimdini Asante-Antwi is research analyst with interest in finance, public policy and social protection. He is also the founder of Project P&L, an organization dedicated to helping financial services companies increase sales on retail loan assets, deposit liabilities and Bancassurance. He is the Executive Director of Tunka Institute for Policy and Program Evaluation (formerly Rural Heights Foundation), a nonprofit committed to improving economic outcomes in rural Ghana.

Email: [email protected]