Multi-product exporters: Learning vs knowing

Sheveleva, Lena ORCID: https://orcid.org/0000-0002-2048-8563 and Krishna, Kala 2013. Multi-product exporters: Learning vs knowing. [Working Paper]. European Trade Study Group 9ETSG). Available at: https://www.etsg.org/ETSG2015/Papers/395.pdf

Preview

PDF - Submitted Pre-Print Version Download (466kB) | Preview

Abstract

It has been documneted across a number of datasets that a large farction of the new products introduced by incumbent exporters are dropped due to low sales within the first year. This regularity is consistent with firms facing uncertaitny in the export market and has spurred the interest in whether firms learn about their market potential as they gain experience in the export market in a Bayesian manner. In the context of multiproduct firms, learning means updating beliefs about a firm-specific effect as the firms observes the success of the products it has exported into the market. Similarly to a number of other studies we document that as much as 40\% of new products by Chinese manufactures to the United States are droped within the first year due to low sales. However, we also document that firms introduce their best products first: products introduced earlier in a firms' exporting career tend to have higher average sales. Guided by these empirical regularities we develop a model where firms learn about the firm-specific brand effect, but unlike earlier models we allow firms to have private knowledge about which of their products are most likely to succeed in the export market. When we take into account such private knowledge we find that learning plays a less important role in export growth than has previously beeen thought.

Item Type:	Monograph (Working Paper)
Status:	Unpublished
Schools:	Schools > Business (Including Economics)
Publisher:	European Trade Study Group 9ETSG)
Last Modified:	02 Jul 2025 14:07
URI:	https://orca.cardiff.ac.uk/id/eprint/177623

Actions (repository staff only)

Edit Item