Stock Prices of Architectural, Engineering, and Construction Firms as Leading Economic Indicator: A Computational Deep-Learning Econometrics Model to Complement the Architecture Billings Index

Rayan H. Assaad, Islam H. El-Adaway

Research output: Contribution to journalArticlepeer-review

5 Scopus citations

Abstract

Investments in the architectural, engineering, and construction (AEC) sector can be used by governments to introduce desired changes in the economy, which gives an indication of the future trends of the economy. In addition, the stock market - as reflected by the Standard and Poor 500 (S&P 500) - creates a sense of confidence about the direction of the US economy. Many studies have shown that stock prices in multiple international industries could be considered as leading economic indicators. Nevertheless, this was not researched within the AEC sector in the United States. This paper addresses this knowledge gap by investigating whether the stock prices of AEC companies in the United States could be used to inform strategic, proactive decision making. First, time series data were collected for the stock prices of 16 large AEC companies as well as for the S&P 500 index. Second, multivariate time series analysis was conducted to examine the type of relationship between the response variable - S&P 500 - and all other input variables. Two econometrics models were developed: a linear, statistical vector autoregression model and a computational, deep learning model based on long-short-term memory (LSTM) networks. Third, the two developed econometrics models were tested on unseen data. Fourth, the best econometrics model was verified using another leading economic indicator which is the Purchasing Managers' Index. The results reflected that the computational LSTM model outperformed the linear statistical model. The findings showed that the stock prices of AEC companies could be considered as an additional and complementing leading economic indicator to the American Institute of Architects' Architecture Billings Index to provide a broader perspective of the business ecosystem. This paper adds to the body of knowledge by providing an econometrics model for helping in optimal investment opportunities, taking informed strategic management and business judgments, and predicting changes in the economy.

Original languageEnglish (US)
Article number04021043
JournalJournal of Architectural Engineering
Volume27
Issue number4
DOIs
StatePublished - Dec 1 2021

All Science Journal Classification (ASJC) codes

  • Architecture
  • Civil and Structural Engineering
  • Building and Construction
  • Visual Arts and Performing Arts

Fingerprint

Dive into the research topics of 'Stock Prices of Architectural, Engineering, and Construction Firms as Leading Economic Indicator: A Computational Deep-Learning Econometrics Model to Complement the Architecture Billings Index'. Together they form a unique fingerprint.

Cite this