Can intangible assets predict future performance? A deep learning approach

• DOI: https://doi.org/10.1108/IJAIM-06-2021-0124
• Published date: 27 October 2021
• Date: 27 October 2021
• Pages: 61-72
• Subject Matter: Accounting & finance,Accounting/accountancy,Accounting methods/systems
• Author: Eleftherios Pechlivanidis,Dimitrios Ginoglou,Panagiotis Barmpoutis

Can intangible assets predict

future performance? A deep

learning approach

Pechlivanidis Eleftherios and Ginoglou Dimitrios

Department of Accounting and Finance,

University of Macedonia, Thessaloniki, Greece, and

Barmpoutis Panagiotis

Department of Computer Science, University College London, London, UK

Abstract

Purpose –The aim of this study is to evaluate of the predictive ability of goodwill and other intangible

assets on forecasting corporate prof‌itability. Subsequently, this study compares the eff‌iciency of deep

learning model to that of other machine learning models such as random forest (RF) and support vector

machine (SVM)as well as traditional statistical methods such as the linear regressionmodel.

Design/methodology/approach –Studies conf‌irm that goodwill and intangibles are valuable assets

that give companies a competitive advantage to increase prof‌itability and shareholders’returns. Thus, by

using as sample Greek-listedf‌inancial data, this study investigates whether or not the inclusion of goodwill

and intangible assets as input variables in this modif‌ied deep learning models contribute to the corporate

prof‌itabilityprediction accuracy. Subsequently, this study comparesthe modif‌ied long-short-term model with

other machinelearning models such as SVMs and RF as well as the traditionalpanel regression model.

Findings –The f‌indings of this paper conf‌irm that goodwill and intangible assets clearly improve the

performance of a deep learning corporate prof‌itability prediction model. Furthermore, this study provides

evidencethat the modif‌ied long short-termmemory model outperforms othermachine learning models suchas

SVMs andRF , as well as traditionalstatistical panel regressionmodel, in predictingcorporate prof‌itability.

Research limitations/implications –Limitation of this study includes the relativelysmall amount of

data available. Furthermore, the aim is to challengethe authors’modif‌ied long short-termmemory by using

listed corporate data of Greece, a code-law country that suffered severely during the recent f‌iscal crisis.

However, this studyproposes that future research may apply deep learningcorporate prof‌itability models on

a bigger pool of data such as STOXX Europe600 companies.

Practical implications –Subsequently, the authors believe that their paper is of interest to different

professional groups, such as f‌inancial analysts and banks, which the authors’paper can support in their

corporate prof‌itability evaluation procedure. Furthermore,as well as shareholders are concerned,this paper

could be of benef‌it in forecasting management’spotential to create future returns. Finally,management may

incorporatethis model in the evaluation processof potential acquisitions of othercompanies.

Originality/value –The contributions of this work can be summarized in the following aspects. This

study providesevidence that by including goodwill and other intangibleassets in the authors’input portfolio,

prediction errors represented by root mean squared error are reduced. A modif‌ied long short-term memory

model is proposed to predict the numerical valueof the prof‌itability (or the prof‌itability ratio) in contrastto

other studieswhich deal with trend predictions, i.e. the binomial outputresult of positive or negative earnings.

Finally, posing an extra challengeto the authors’deep learning model, the authors’usedf‌inancial statements

according to International Financial Reporting Standard data of listed companies in Greece, a code-law

country that suffered during the recent f‌iscal debt crisis, heavily inf‌luenced by tax legislation and

characterizedby its lower investors’protection compared to common-lawcountries.

Keywords Corporate prof‌itability, Goodwill, Intangible assets, Deep learning model,

Long short-term memory networks

Paper type Research paper

Intangible

assets

61

Received15 June 2021

Revised7 September 2021

Accepted12 October 2021

InternationalJournal of

Accounting& Information

Management

Vol.30 No. 1, 2022

pp. 61-72

© Emerald Publishing Limited

1834-7649

DOI 10.1108/IJAIM-06-2021-0124

The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1834-7649.htm