Processes and Accuracy of Cash Flow Forecasting: A Case Study of a Multinational Corporation

• DOI: http://doi.org/10.1111/jacf.12301
• Date: 01 June 2018
• Author: Martin Glaum,Kati Schnürer,Peter Schmidt
• Published date: 01 June 2018

In This Issue: Notes from the Field

Investors as Stewards of the Commons? 8George Serafeim, Harvard Business School

Rethinking the Purpose of the Corporation 18 Edward J. Waitzer, Stikeman Elliott LLP

The ESG Integration Paradox 22 Michael Cappucci, Harvard Management Company

Building a Bridge between Marketing and Finance 29 Ryan Barker, BERA Brand Management, and Greg Milano,

Fortuna Advisors

“Big Data” Analysis: Putting the Data Cart Before the Modelling Horse? 40 Graham D. Barr, Theodor J. Stewart & Brian S. Kantor,

University of Cape Town

Debt Crisis Looming? Yes, Corporate Debt Expanded

but Don’t Panic Over the Prospect of BBB Downgrades

45 Martin Fridson, Lehmann Livian Fridson Advisors

Buyout Transactions in the German-speaking Region:

Determinants of Abnormal Performance and Unlevered Returns

50 Fabian Söffge and Reiner Braun, Technical University

of Munich

Processes and Accuracy of Cash Flow Forecasting:

A Case Study of a Multinational Corporation

65 Martin Glaum, WHU - Otto Beisheim School of Management,

Peter Schmidt, Justus-Liebig-Universität Giessen, and

Kati Schnürer, Bayer AG & Justus-Liebig-Universität Giessen

An Empirical Study of Insurance Performance Measure 83 Sai Ranjani Bharathkumar, XLRI Jamshedpur, India

Valuation of Corporate Innovation and the Pricing of Risk in the

Biopharmaceutical Industry: The Case of Gilead

92 Richard Ebil Ottoo, Global Association of Risk Professionals

(GARP)

VOLUME 30 | NUMBER 2 | SPRING/SUMMER 2018

A P PLIED COR P O R ATE FINANC E

Journal of

Journal of Applied Corporate Finance • Volume 30 Number 2 Spring 2018 65

Processes and Accuracy of Cash Flow Forecasting:

A Case Study of a Multinational Corporation

*We thank Niamh Brennan, Alexander Burck and Tom O’Brien for helpful comments

to earlier versions of the paper. We are also grateful for comments and suggestions re-

ceived at the EAA Annual Conference 2017 and from workshop participants at Justus-

Liebig-Universität Giessen, ESSEC, Université Toulouse Capitole, Universität St. Gallen

and Wirtschaftsuniversität Wien.

**One of the authors (Kati Schnürer) was pursuing research for her PhD while being

employed at Bayer AG’s central nance department.

1. Martin Glaum, Peter Schmidt, and Kati Schnürer, “What Determines Managers’

Perceptions of Cash Flow Forecasting Quality? Evidence From a Multinational Corpora-

tion,” Journal of International Financial Management and Accounting, 27(3), 298–

346, 2016.

2. In line with nancial reporting requirements, Bayer’s reporting systems are de-

signed to record not cash inows and outows but income and expenses. Our quantita-

tive analyses are therefore based on forecasted and booked invoices, to approximate the

related cash ows.

3. This nding is consistent with the perception of the interview participants. All 20

managers interviewed were of the opinion that the forecasting of cash outows is more

challenging than the forecasting of cash inows.

ash ow forecasting is a ke y element of compa-

nies’ nancial ma nagement. Short-term cash

flow forecast s are a prerequisite for liquidity

management and hedging of f inancial risk s,

and longer-term cash ow forecasts are t he basis for invest-

ment and nancing decisions. However, despite its pivotal

importance, cas h ow forecasting is ra rely considered in the

academic literature, presum ably because researchers normally

have no access to company-internal processes a nd data.1

In this study, we explain how cash ow foreca sting is

organized at Bayer, a large multinationa l company headquar-

tered in Germany, and which factors inuence the accuracy

of its forecasts. Our rese arch focuses on ca sh ow forecasts

based on the direct met hod. Within our sample company,

such forecasts are prepared t hree times a year with horizons

of twelve months. Each time, th is involves generating about

62,000 individua l forecasting items. ese forecasts form the

basis of the company’s liquidity and nancia l risk manage-

ment, in particula r, its foreign exchange ris k hedging. We rst

investigate how local man agers in Bayer’s entities across the

world derive the forecasts, i.e., what information they use a s

input, how they validate it, and how they dea l with potential

bias caused by mana gerial incentive systems. We also analy ze

whether forecasting proce sses are aected by charac teristics of

the local entities, such as bu siness area, size, region, or specic

local conditions, and ulti mately whether forecasting practices

and entity characteri stics aect forecast accurac y.

Our study is based on 20 semi-structured interviews

conducted in 2015 with managers of Bayer who are respon-

sible for cash ow forecasting, and on participant observation.

To ensure that all major types of business transactions are

included in the sample, we selected managers of the company’s

most important legal entities. We structure the information

obtained from the interviews by the steps in the forecasting

process, by cash ow components, and by entity characteristics.

We carefully examine the responses for each component and

analyze whether forecasting processes dier across cash ow

components or across the entities’ characteristics. We then use

a large company-internal dataset of cash ow forecasts and

cash ow realizations for the participating entities to calculate

weighted absolute percentage forecasting errors (WAPE).2 We

use this measure to analyze the association between forecasting

procedures and legal entity characteristics on the one hand, and

forecast accuracy on the other hand.

Our ndings reveal that cash ow forecasting procedures

vary substantially across our sample company’s entities. While

the central nance department gives general guidance on the

required cash ow forecasting output and provides direction

on the input to be used, there are no detailed instructions on

how forecasts are to be prepared. Instead, local managers are

free to gear their forecasting practices to the dynamics, the

uncertainty, and the complexity of their entities’ businesses,

to local nancial market conditions, and to their respective

organizational structures. As a consequence, they use dierent

forecasting inputs and validate forecasting inputs and output

with dierent intensities, and they also dier in how they treat

possible biases in input data. ese ndings document the

limits of standardization and central control in large multi-

national corporations resulting from local managers’ need for

exibility to cope with the heterogeneity and dynamism of

their environments. At the same time, however, local dier-

entiation increases complexity and may increase errors.

Our quantitative analy sis of forecasting errors fu rther

reveals that foreca sts of receipts from customers (cash inows)

are more accurate than forec asts of payments to suppliers (cash

out ows). 3 Moreover, foreca sting pra ctices aect f orecast

accuracy. Outow forecasts a re more accurate if mana gers

intensively validate forecast ing input, whereas inow forecasts

are more accurate t hey eliminate input biases that may result

from internal target set ting or from other managerial incen-

tives, and if they ca refully validate their forecasti ng output.

An extensive academic literature exists on discounted cash

C

by Martin Glaum, WHU - Otto Beisheim School of Management, Peter Schmidt, Justus-Liebig-Universität

Giessen, and Kati Schnürer, Bayer AG & Justus-Liebig-Universität Giessen*, **