Contributions in Merlin

null

The high failure rate among new business ventures is usually chalked up to the fundamental uncertainty of the process. In actuality, say McGrath and Keil, flawed ways of assessing and managing ventures may account for the disappointing amount of value they generate. Instead of taking the go/no-go approach, whereby a project either advances toward launch or is killed, decision makers should consider a range of alternatives: recycling the venture by aiming it at a new target market; spinning it off to other owners or a joint venture; spinning it in to an established business unit; or salvaging useful elements such as technologies, capabilities, knowledge, and patents. Firms that excel in value extraction, the "value captors," whose practices and mind-set this article explores have created formal processes to systematically mine successes, failures, and everything in between. They know that a venture should be treated like a scientific experiment, in which learning plays a critical role. They are ready to seize new opportunities if a venture falters on its original course. They foster networks to promote cooperation and collaboration between established business leaders and venture teams and involve people from throughout the company in the venture review process. They don't allow financial criteria to dominate the reviews, and they recognize that the best people to launch a business may not be the ones who developed the idea. If your innovation pipeline is dry, your promising projects are being strangled for lack of a speedy payback, or someone else has made a fabulous business out of a slightly altered idea that you abandoned, consider the value captor's path. INSETS: Ten Telltale Signs of a Flawed Venturing Process;The Value Captor's Process;How Texas Instruments Discovered a Promising Business.

null

This paper introduces a new filter, called least-squares (LS) filter, as an alternative to the widely used Hodrick-Prescott (HP) filter. The LS filter has two properties which distinguish it from the HP filter. First, the LS filter rather than the HP filter uses first differences as a measure of the smoothness of a nowhere differentiable random variable. Second, the LS filter rather than the HP filter uses a reference time series to avoid the subjective choice of weights.