Essay 2018 06 06 000EOZ | EssaySauce.com

Effect of negative interest rates on corporate cash positions

June 2018

Robert-Jan Terpstra

Table of Contents

Executive Summary

In this study the effects of negative interest rates and the key determinants of corporate cash balances were investigated. 444 listed companies in the Eurozone over the period 01/01/2013-31/12/2017 were examined. A regression analysis was performed with cash as the dependent variable and Interest, Market-to-book ratio, Cash Flow, Net working Capital, Leverage, Firm size and Dividend as the independent variables.

It was found that Market-to-book ratio and Net working Capital had a significant positive impact on cash positions. Firm size and dividend were found to both have significant negative impact on cash positions.

The study didn’t find any impact of interest turning negative on corporate cash balances. However, as firms need to update their policies in order to change the way in which they invest their cash, the effects might only become visible after several quarters. In addition, operational deposits have actually become more interesting for banks under Basel 3, which is an effect that wasn’t adjusted for in this study.

Table of contents

Executive Summary 2

Introduction 4

Literature review and research hypothesis 5

Methodology 9

Results 11

Conclusion 13

Appendix 14

Literature 18

Introduction

In recent years we have seen the ECB push interest rates in the Eurozone lower and lower. The Euro Interbank Offered Rate ‘Euribor,’ which is the average rate at which Eurozone banks offer to lend unsecured funds to other banks in the interbank market, (Wikipedia) has been moving steadily lower on the back of this as can be seen in graph 1. As the interest rate at which Eurozone banks are lending money to one another has turned negative, we would expect this to have had an effect on deposit rates for their corporate clients as well.

The question is what this interest effect has had on the size of corporate cash balances. Banks and consultants have come up with a range of solutions to avoid the immediate effects, but still they haven’t been able to fully offset the effects. As Zanders also mentions in their paper ‘How negative interest rates will affect treasury’ from a psychological perspective there is a difference between receiving a (small) positive interest on cash balances or having to pay a negative interest. So far we haven’t come across any research on the effects of negative yields on corporate cash balances so therefore we believe it would be interesting to see whether interest has become a factor influencing cash balances now that rates have turned negative.

Graph 1: 1 MONTH Euribor fixing, Source: Reuters Eikon

Literature review and research hypothesis

There is a limited amount of research written about the key determinants of cash holdings by corporates. Cash holdings are usually explained by three theories: trade-off theory, pecking order theory and the free cash flow theory. (Ferreira & Vilela, 2004)

The trade-off theory weighs the benefits of debt that result from shielding cash flows from taxes against the costs financial distress associated with leverage. (Berk, De Marzo) The level of cash held is based on the marginal benefit and cost of holding cash, as Liquid holdings provide a low yield. (Ali & Yousaf, 2013)

According to the pecking order theory (Myers & Maljuf, 1984) managers prefer to use retained earnings before issuing debt and only use equity issuance as a last option to minimize asymmetric information costs and other financing costs.

The free cash flow theory developed by Jensen (1986) states that managers will use free cash flow to invest in projects with negative NPV instead of returning cash to shareholders. Cash increases the agency problem as managers have the opportunity to invest in projects that may not be in the shareholders best interest.

In perfect capital markets holding liquid assets would have no added value so frictions have to be able to explain why external financing isn’t fully replaceable by internal finance. Opler et al (1999) examined the factors influencing the holdings of cash and marketable securities by listed US firms in the period 1971-1994. Their findings suggest that there are two reasons for holding cash positions: the transaction cost motive and the precautionary motive.

The transaction costs motive has to do with the costs of raising funds regardless whether this is done by accessing capital markets or by selling assets which motivates companies to raise funds on an irregular basis and to keep cash and marketable securities as a buffer.

The precautionary motive is being afraid of not having sufficient liquid assets available to invest in profitable projects. Firms might prefer to hold cash to avoid financial distress. According to Drobetz & Gr”ninger (2007) it is caused by information asymmetries, agency costs and the opportunity costs of foregone investments. Opler et al (1999) find that firms with strong investment opportunities have a tendency to hold more cash as the potential costs of insufficient liquidity would be larger.

Brav et al (2005) conducted a survey with 384 CFOs and treasurers and found that managers are very reluctant to cut dividends, that dividends are smoothed through time and that dividend is linked to long term sustainable earnings. Companies now often use buy backs as a way of returning cash to investors while remaining flexible. This would imply that companies that pay dividend would maintain higher cash balances in order to make sure they have sufficient means to maintain a stable program. On the other hand external monitoring could be higher for firms that pay dividend as argued by Drobetz & Gr”ninger (2007), which would make it easier for them to attract new capital.

Drobetz & Gr”ninger (2007) their results indicate that firms with higher leverage hold less cash which is in support of the idea that the opportunity costs of holding cash rises with leverage. They also find that large firms hold less cash due to economies of scale in capital raisings. There seems to be a positive relationship between cash flows and cash holdings. A significant relationship between market-to-book and cash holdings was not found.

Ali et al (2013) performed a study on German companies in the period 2000-2010. They find that cash holdings are negatively correlated to the size of firm, leverage, working capital and market to book value. This last factor is used as a proxy for growth opportunities. Saddour (2006) describes that one would expect a positive relationship between growth opportunities and cash balances as firms need to make sure they have sufficient funds to seize investment opportunities. In addition, companies with strong growth opportunities have greater financial distress costs.

The negative relationship between cash and working capital Ali et al (2013) find is in line with the theory that liquid assets, for which working capital is a proxy, can easily be converted in cash and therefore is a substitute for cash balances.

Saddour (2006) investigated the determinants of cash holdings for French firms over the period 1998-2002. His results point out that French firms increase their cash level when cash flow level is high and reduce it when leverage increases. Cash flow can be used as a source of liquidity to finance investments so he would have expected cash flow to be a substitute for cash holdings and therefore cause a negative relationship between cash flow and cash balances.

There seems to be little written about the effects of interest rates on cash balances. In recent years we have witnessed short term EUR interest rates move into negative territory. Assuming banks have passed on the negative interest rates to their (large) corporate clients and have started charging negative rates on their EUR cash deposits, it would be interesting to test whether this has had an effect on cash balances held.

On the other hand one could argue that corporates look at the overall impact on ROE and in that case whether they receive 10bp, 0 bp or -10 bp on a deposit only has minimal impact if they want to achieve an overall ROE of for example 10%. The 2017 JP Morgan Asset management Global liquidity Investment Survey which has 378 global responses across different sectors and sizes find that ‘a large majority of respondents are considering policy changes to allow increased credit risk, more interest rate risk and the use of currency swaps.’ The survey has found that since 2015 35% of European participants to the survey, and 30% of firms with >USD 5 bn in cash assets, have actually changed their policies to allow for more credit risk. This indicates that behaviour on the back of the negative short term interest rates is shifting in order to avoid the effects of negative interest rates.

The treasury consultancy firm Zanders writes in its piece ‘How negative interest rates will affect treasury’ that ‘from a psychological point of view, it is difficult for people to accept a low or even negative yield on investments. This aversion to loss can result in increased risk taking.’ Therefore a strong treasury framework should be in place to make sure the organization continues to make rational investment decisions. Zanders provides a few ideas in order to reduce the effects of negative rates:

Centralize cash to make sure only the header account of a cash pool is affected by the negative rates

Corporates can place cash on a time deposit with higher rates

‘Invest’ excess cash in a supplier finance program which can lead to discounts, enhanced relationships with suppliers and reduced risk of default of suppliers as their working capital is reduced.

Adam Hayter of RBS says in the article ‘Negative rates: implications for banks and corporates’ that firms should look at their cash forecasting to see if any efficiencies can be made to cope with negative rates. ‘Using day-to-day, operational bank balances more efficiently and putting more money into long-term vehicles with rates above zero would certainly help their situation.

In the article ‘Finding return in a Sub-zero world’ head of EMEA liquidity at Bank of America Merrill Lynch ‘Suzanne Janse van Renshoven recommends firms to look at a range of options including ‘interest optimisation, self-funding of supply chain finance, and early execution of existing business strategy.

Azara et al (2015) find that when in the US restrictions for firms to invest their cash reserves in interests bearing accounts was dropped, which reduces the costs associated with holding large cash reserves, that companies increased their cash reserves. The question obviously is whether lowering yields in the Eurozone will have the opposite effect.

In the FT article ‘German companies hit by negative interest rates’ an IFO study on German companies is cited which finds that almost 20% of all surveyed companies have had their banks trying to charge negative interest rates. As many switched banks or moved assets only one in ten eventually was charged negative rates. According to the article ‘About half of companies threatened with the charges began to negotiate with the bank, while more than one-third joined another lender. Thirty per cent bought other assets or repaid loans and 29 per cent redeployed their money elsewhere in their companies.’

Greenwhich Associates has published a report called ‘Amid Negative Rates and Bank Turmoil, Cash Management Emerges as Key Challenge for European Companies’ in which they describe the findings of interviews with 2585 European corporates. They find that from 2014 to 2015 corporate cash balances held by banks and cash management providers declined from 69% to 67%. In addition the share of these balances used to fund M&A activities increased from 2% to 3% and the amount used to pay shareholders rose to % from 2%. Greenwhich Associates Director Mr Miarka says: ‘We expect these trends to continue and for more companies to direct cash holdings to more productive uses like dividends, share buybacks and M&A for as long as negative interest rates persist, (..) We also expect companies to be more receptive to overtures from non-bank providers that can help them make better use of surplus cash.’

The German paper Handelsblatt Global reports in its article ‘Below Zero Freezes Out German Companies’ that companies are spreading their money around different banks in order to remain below the threshold above which they are charged negative rates. They also provide anecdotal evidence on a company which is investing its surplus cash for 3 and 6 months in order to avoid negative rates.

Based on the available literature there are indications that negative interest rates are having an effect on the size of corporate cash balances. The main hypothesis this study therefore aims to test is the following:

Null hypothesis (H0): Negative rates in the Eurozone are not causing a decrease in corporate cash balances

Alternate hypothesis (H1): Negative rates in the Eurozone are causing a decrease in corporate cash balances

Methodology

This study was carried out by analysing the quarterly data of 444 listed non-financial companies in the Eurozone over the period 01/01/2013 -31/12/2017. The minimum market cap for qualifying was 500m EUR. Financial institutions and companies not reporting on a quarterly basis were excluded from the study. The data was sourced from Bloomberg.

Participating companies had to be listed in the years 2014 and 2015 as 1month Euribor turned negative in 2015 and the effect had to be incorporated in the analysis.

By performing a regression analysis with ‘cash / (book value of assets ‘ cash and cash equivalents)’ as the dependent variable and the below variables as independent factors the key determinants that have an effect on cash balances were investigated. In order to build on previous research, all the variables used which could have an effect on the size of cash balances, were based on the research of Gill & Shah (2012) for Canada. The variables ‘Board Size’, ‘CEO duality’ and ‘Industry’ were deleted and ‘Interest’ was added.

Independent variables:

Interest = dummy based on 1mth Euribor which takes a value of ‘1’ if level is below zero and ‘0’ if level is above zero. ‘Euribor’ is the average rate at which Eurozone banks offer to lend unsecured funds to other banks in the interbank market. In this study it is assumed that 1 month Euribor is a good proxy for the interest banks pay on corporate cash balances as banks are expected to pass on the negative rates to their clients.

MTB (Market-to-book-ratio for firm i in time t) = (Book value of Assets ‘ Book value of equity + Market value of Equity) / Book value of assets. MTB is used as a proxy for future growth prospects of the different companies in our sample.

CF = (Cash flow to net asset ratio for firm i in time t) = (after tax profit + depreciation) / (Book value of Assets ‘ Cash and cash equivalents)

NWC (Net working capital to asset ratio for firm i in time t) = (net current assets ‘ Cash and cash equivalents) / (Book value of Assets ‘ Cash and cash equivalents)

Leverage (Leverage for firm i in time t) = Total (gross) debt / (Book value of Assets ‘ Cash and cash equivalents)

Firm size (Firm size of firm i in time t) = Natural log of Book value of assets of firm

Dividend dummy (Dividends paid by firm i in time t) = ‘1’is used if firm paid dividends each year otherwise ‘0’

” = error term

Dependent variable

CASH (Corporate cash holdings for firm i in time t) = Cash and cash equivalents / (Book value of assets ‘ Cash and cash equivalents)

The regression model is as follows:

CASH = a + b * Interest + c * MTB + d * CF + e * NWC + f * Leverage + g * Firm size + h * Dividend dummy + ”

Results

The study contains 8270 observations as can be seen in Table 1. By winsorizing at the 1st and 99th percentiles the data was corrected for outliers. The (adjusted) R2 of the linear regression that was performed came in at 0.1999 which implies that about 20% of the variance in corporate cash holdings can be explained by the model. Also the ANOVA table shows the regression is highly significant at 0.000.

All Independent variables apart from Interest appear highly significant (at 0.000 level) based on the results in table 3. However after adjusting for clustering of the standard error for both period and company the significance of the coefficients changes considerably. As can be found in table 4, NWC is significant at the 0% level, Dividend at 0.1%, MTB at 1% and Firmsize at 10%. There doesn’t seem to be an issue with multicollinearity as the Variance Inflation Factor (VIF) of the variables are around 1. Neither do the correlation levels point to any issues with multicollinearity.

The positive NWC coefficient we find, significant at the 0% level, is not in line with the results of Ferreira et al (2004) who find a negative coefficient and our results are not in line with the trade-off model that indicates that liquid assets act as substitutes for cash. Also Ali et al (2013) find a different result for the German market and Gill & Shah (2012) for Canada.

Dividend appears to be negatively affecting cash positions which is significant at 0.1%. This is different from what Drobetz & Gr”ninger (2007) found on Switzerland but would be in line with the idea that external monitoring could be higher for firms that pay dividend which would make it easier for them to attract new capital.

MTB appears to be significant at the 1% level which is in line with what the trade-off and pecking order theories prescribe, companies with better investment opportunities have larger cash positions as they need to make sure they have sufficient funds available to grasp interesting investment opportunities.

Firm size has a negative coefficient but is only significant at 10%. This finding follows the suggestion that larger firms face lower costs when attracting funds and is in line with the results from Ferreira & Vilela (2004) on listed companies in EMU countries, Saddour (2006) on France and Drobetz & Gr”ninger (2006) on Switzerland and finally Opler et al (1997) on US firms.

CF, Interest and Leverage don’t seem to have a significant impact on cash positions held by corporates. Most research so far has found evidence that there is a relationship between cash flow and cash balances and between leverage and cash positions, our results however didn’t find this.

The study didn’t find any evidence that negative interest rates have had an impact on the size of cash being held by European corporates. This leads to the conclusion that hypothesis H0 cannot be rejected on the basis of the results of this study.

The problem might be that we were looking for an immediate impact of negative interest rates on the cash balances of corporates. The question is whether banks have started passing on negative rates to their corporate clients after the ECB moved their deposit rates into negative territory or whether there has been a delay. Under Basel 3 banks are forced to hold sufficient highly liquid assets to cover its liquidity needs over a 30 day period which is tested via the following formula: (Lekatis, 2015)

LCR=(Stock of high quality liquid assets)/(Net cash outflows over a 30 day time period)’100%

Deposits have to be taken into account when calculating the net cash outflows in times of stress. Therefore high quality liquid assets have to be held to cover for potential outflows. For operational deposits, which consist of general working capital and cash held for transactional purposes, a bank has to hold liquid assets equal to 25% of the size of the deposit, where for non-operational deposits this percentage is equal to 40%. (Citi Academy) The reason for this difference is that ‘Basel III considers operational cash to be ‘stickier’ than non-operational cash, in the sense that companies are likely to maintain operational deposits with their primary transaction bank(s).’ Our study wasn’t able to distinguish between operational and non-operational deposits which might be influencing the visibility of the effects. In addition there might have been some delay in adjusting the ‘investment policy’ so the actual effects might only become visible after several quarters. As changing the tenor or type of cash investments might cause these to no longer qualify as ‘cash and cash equivalents’

Conclusion

This study wasn’t able to find any evidence of negative interest rates having an impact on corporate cash balances. However, as firms need time to update their policies and systems in order to change the way in which they invest their cash, the effects might only become visible after several quarters. In addition, operational deposits have actually become more interesting for banks under Basel 3, which is an effect that wasn’t adjusted for in this study.

It was found that Market-to-book ratio and Net working Capital had a significant positive impact on cash positions. The positive impact of MTB is in line with the results of previous studies and the trade-off and pecking order theories. Companies with better investment opportunities have larger cash positions as they need to make sure they have sufficient funds available to grasp interesting investment opportunities. The positive NWC coefficient we find, doesn’t follow the results of other studies who find a negative coefficient and our results are not in line with the trade-off model that indicates that liquid assets act as substitutes for cash.

Firm size and dividend were found to both have significant negative impact on cash positions. Dividend’s negative effect is different from what Drobetz & Gr”ninger (2007) found on Switzerland but would be in line with the idea that external monitoring could be higher for firms that pay dividend which would make it easier for them to attract new capital. Our finding on Firm size is only significant at 10%, but follows the findings of other studies and the suggestion that larger firms face lower costs when attracting funds.

CF, Interest and Leverage don’t seem to have a significant impact on cash positions held by corporates. Most research done so far, has found evidence that there is a relationship between cash flow and cash balances and between leverage and cash positions, our results however didn’t find this.

Appendix

Table 1 Descriptive Statistics

N Minimum Maximum Mean Std. Deviation

cash 8270 .0024 1.4297 .138094 .1974688

interest 8270 0 1 .57 .495

MTB 8270 .6026 6.9092 1.717948 1.0980262

CF 8270 -.1249 .1166 .023445 .0288091

NWC 8270 .0063 .9190 .348762 .2099467

Leverage 8270 .0000 53.1400 1.570828 6.4032284

Firm size 8270 4.2700 12.2538 8.025949 1.6294941

Dividend dummy 8270 0 1 .24 .427

Valid N (listwise) 8270

Table 2 ANOVAa

Model Sum of Squares df Mean Square F Sig.

1 Regression 64.668 7 9.238 296.098 .000b

Residual 257.773 8262 .031

Total 322.441 8269

a. Dependent Variable: Cash

b. Predictors: (Constant), Dividend dummy, MTB, Interest, Leverage, NWC, CF, Firm size

Table 3 Regression results before adjusting for clustering of the standard error

lm(formula = regr, data = dat)

Residuals:

Min 1Q Median 3Q Max

-0.81039 -0.07888 -0.02839 0.02982 1.40911

Coefficients:

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.1775749 0.0132781 13.373 < 2e-16 ***

Interest -0.0032239 0.0039521 -0.816 0.414672

MTB 0.0520955 0.0020293 25.672 < 2e-16 ***

CF -0.8449798 0.0712560 -11.858 < 2e-16 ***

NWC 0.0719593 0.0096037 7.493 7.43e-14 ***

Leverage 0.0054225 0.0003124 17.355 < 2e-16 ***

Firmsize -0.0170572 0.0013445 -12.686 < 2e-16 ***

Dividenddummy -0.0168279 0.0046017 -3.657 0.000257 ***

—

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ‘ 1

Residual standard error: 0.1766 on 8262 degrees of freedom

Multiple R-squared: 0.2006, Adjusted R-squared: 0.1999

F-statistic: 296.1 on 7 and 8262 DF, p-value: < 2.2e-16

Table 4 Regression results after adjusting for clustering based on Company and Quarter

t test of coefficients:

Estimate Std. Error t value Pr(>|t|)

(Intercept) 0.1775749 0.0871525 2.0375 0.041630 *

Interest -0.0032239 0.0077081 -0.4183 0.675774

MTB 0.0520955 0.0244654 2.1293 0.033255 *

CF -0.8449798 0.7501479 -1.1264 0.260022

NWC 0.0719593 0.0154319 4.6630 3.165e-06 ***

Leverage 0.0054225 0.0033128 1.6368 0.101700

Firmsize -0.0170572 0.0099190 -1.7197 0.085533 .

Dividenddummy -0.0168279 0.0056198 -2.9944 0.002758 **

—

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ‘ 1

Table 5 Test results on multicolinearity

Variables VIF

1 cash 1.250152

2 interest 1.013628

3 MTB 1.436126

4 CF 1.157616

5 NWC 1.086353

6 Leverage 1.115371

7 Firmsize 1.296509

8 Dividenddummy 1.024271

Table 6 Pearson correlation results

Cash Interest MTB CF NWC Leverage Firm size Dividend dummy

Cash Pearson Correlation 1 -.017 .337** -.026* .165** .241** -.300** -.062**

Sig. (2-tailed) .113 .000 .018 .000 .000 .000 .000