Determinants and usefulness of analysts' cash flow forecasts: evidence from Australia

• Pages: 4-21
• Published date: 22 February 2013
• DOI: https://doi.org/10.1108/18347641311299722
• Date: 22 February 2013
• Author: Kamran Ahmed,Muhammad Jahangir Ali
• Subject Matter: Accounting & finance

Determinants and usefulness

of analysts’ cash f‌low forecasts:

evidence from Australia

Kamran Ahmed and Muhammad Jahangir Ali

Department of Accounting, La Trobe University, Melbourne, Australia

Abstract

Purpose – The purpose of this paper is to examine the determinants of analysts’ operating cash f‌low

forecasts of Australian listed f‌irms and whether or not such forecasts improve the usefulness of

earnings and predictive ability of current cash f‌lows.

Design/methodology/approach – The authors used a large sample of f‌irms for which both cash

f‌lows and earnings forecasts were available over a period between 1993 and 2003, and employed both

univariate and logistic regression analyses.

Findings – It wasfound that analysts forecast both operating cash f‌lows and earnings when the f‌irms

are more complex in operations and when the size of the f‌irm is relatively small. Further, it was found

that cash f‌low forecasts improve the usefulness of earnings and predictive ability of current cash f‌lows.

Originality/value – This study contributes to current understanding of analysts’ forecast behaviour

regarding dissemination of operating cash f‌low information and usefulness of cash f‌low forecasts.

Keywords Cash f‌low, Earnings,Financial forecasting, Cash f‌low forecasts,Earnings forecasts,

Analysts’ forecasts, Usefulness of earnings,Predictive ability of operating cashf‌lows

Paper type Research paper

This study examines the determinants of cash f‌low forecasts by analysts for

Australian listed f‌irms and whether or not such determinants are important

for investment decisions. During the last two decades, it has become increasingly

common for analysts to forecast cash f‌lows from operations (CFO) along with earnings.

Earnings forecasts by f‌irms’ management and f‌inancial analysts are very important

for the valuation of f‌irms and f‌inancial reporting quality as they relate to realized and

expected earnings. Analysts are prominent information intermediaries in capital

markets because they engage in private information research, perform prospective

analysis aimed at forecasting a f‌irm’s future earnings and cash f‌lows, and conduct

retrospective analysis that interprets past events (Beaver et al., 1980). The importance

of analysts’ role in capital markets has spurred research showing that analysts

inf‌luence the informational eff‌iciency of capital markets (Frankel et al., 2006).

Prior research, including that of O’Brien (1988) and Kothari (2001), has used

analysts’ earnings forecasts as proxies for the unobservable market expectations

of future earnings realization. Although earnings are a summary of performance

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1834-7649.htm

Accepting editor: Lee J. Yao

The authors would like to acknowledge f‌inancial supports from La Trobe University and

AFAANZ, and comments and suggestions from the participants at the Department of

Accounting seminar, La Trobe University and the Asian-Pacif‌ic Conference on International

Accounting Issues in Paris 2008. Thanks are also due to two anonymous referees for their

constructive comments. The authors take responsibility for any remaining errors.

IJAIM

21,1

4

Received 19 January 2011

Accepted 15 March 2011

International Journal of Accounting

and Information Management

Vol. 21 No. 1, 2013

pp. 4-21

qEmerald Group Publishing Limited

1834-7649

DOI 10.1108/18347641311299722

measures that comprise cash f‌lows and accruals, they explain only a small fraction of

the total variation in share performance because earnings may not ref‌lect the

underlying economic events in a timely manner (Hayn, 1995). Consequently, recent

literature emphasises both cash f‌low forecasts and earnings forecasts as important for

f‌irm valuation and performance measure purposes (DeFond and Hung, 2003). There is

evidence that a large portion of the investment community relies more on cash f‌lows

than earnings in the decision-making processes (FASB, 1978; Golub and

Huffman, 1984; Call, 2008).

Investors can view the presence of a cash f‌low forecast as a sign of the importance

of the f‌irm’s underlying cash f‌low information (Call, 2008). Call f‌inds that investors in

the US put more emphasis on cash f‌low information into stock prices when such

information is available. Prior to that McInnis and Collins (2006) f‌ind that cash f‌low

forecasts increase transparency of accrual manipulations used to meet earnings targets

and that f‌irms with cash f‌low forecasts have higher accrual quality. They also f‌ind that

analysts’ cash f‌low forecasts provide an effective earnings management constraint.

Although it has been noted that f‌inancial analysts provide earnings per share (EPS)

forecasts for some f‌irms and EPS and operating cash f‌lows (CFO) forecasts for other

f‌irms, there is very limited empirical evidence on why analysts treat f‌irms differently

in forecasting and whether investors in markets outside the USA value such CFO

forecasts. This study is an attempt to understand analysts’ forecasting behaviour with

respect to a small but eff‌icient capital market such as Australia.

Australiahas a highly developed capitalmarket, where f‌irms reach outto secure equity

and debt f‌inancefrom international and domesticinvestors. Further,the Australian Stock

Exchange (ASX) is considerably smaller than stock exchanges in the USA or the UK.

The ASX has fewer listed companies with concentration of stock exchange value,

volume trading among larger companies and institutional investments in Australia

focusingon larger companies. Those includedin market-representativeindices that result

in reduction in the overallbreadth of analyst following and high information asymmetry

(Fleminget al., 2005). It is thereforeargued that analysts would beinclined to provide cash

f‌low information aboutf‌irms to reduce information asymmetry.

Our empirical analysis consists of two parts:

(1) to examine the f‌irm-specif‌ic attributes associated with forecasting of cash f‌lows; and

(2) to assess the extent to which such cash f‌low forecasts are useful to equity

investors.

To this end, we examined all available annual earnings forecasts for Australian f‌irms

from the Institutional Brokers’ Estimate System (I/B/E/S) database from 1993 to 2003

and developed models on the basis of accounting, operating and f‌inancing

characteristics of f‌irms that might inf‌luence cash f‌low forecasts, and their usefulness

in interpreting and predicting earnings.

The results, based on a large sample comprising 2,656 f‌irm-year observations over a

period of 11 years, show that complexity of operations as measured by number of

geographic segments is positively related to earnings and cash f‌low forecasts and that

size is negatively associated with earnings and cash f‌low forecasts. This implies that

analysts disseminate operating cash f‌lows along with earnings information for f‌irms

that are more complex while they forecast only earnings information for large

Australian f‌irms. We also f‌ind that cash f‌low forecasts have an incremental effect on

Analysts’ cash

f‌low forecasts

5