This analysis of the Intrepid Small Cap Fund includes a performance attribution report which provides a step-by-step analysis framework, detailed charts, and descriptions related to Mr. Cinnamond’s selection and timing skills over the past 18-months. Several surprising observations emerged, most specifically regarding the incremental “timing contribution” figures given Mr. Cinnamond’s sizeable moves into cash and other sectors over the period of high market volatility. Most of the excess performance, as determined by the sector specific returns-based style analysis model, was generated through “selection”.

Introduction

The Intrepid Small Cap Fund, managed by Eric Cinnamond of Intrepid Capital Funds, has garnered noteworthy media attention in 2009. Bloomberg Online’s June 3, 2009 article featured Cinnamond’s fund as the only diversified stock manager to outperform Bill Gross’ venerable Pimco Total Return Bond fund over the trailing 3-year period (through 5/26/09). Wall Street Journal’s October 5, 2009 article puts this fund in second place in the Winners’ Circle contest for the 12-months through September 2009 (posting a return of 29%). What explains Intrepid Small Cap’s outperformance? MPI takes a closer look using daily data and returns-based style analysis (RBSA), and finds Mr. Cinnamond’s selection skills over the past nineteen months have been strikingly strong relative to other small capitalization mutual funds.

We illustrate a step-by-step performance attribution approach using returns-based style analysis, daily data, and holdings information. High frequency daily data enables practitioners to detect style changes with greater speed and accuracy. Daily data also provides a better performance attribution especially for a fund with short history and high turnover. The report contains more detailed descriptions along with supporting charts and tables generated by an mpi Stylus Pro performance attribution template (note that this template is available for client use; please contact Support at support@markovprocesses.com for more information). The report demonstrates how a performance attribution narrative can be crafted in a straightforward, efficient and compelling way.

Background

Mr. Cinnamond is a devout bottom-up stock picker who is openly resistant to efficient market and asset allocation theories and practices. Although Intrepid Small Cap is categorized as a small cap value manager, Mr. Cinnamond’s strict focus on small capitalization companies’ operating fundamentals can be described as being somewhat style-agnostic. He focuses on businesses with strong free cash flow, asset and transaction valuations, respectively. According to the Wall Street Journal’s online “Ask a Fund Manager” series in April 2009, Mr. Cinnamond typically has a potential buy list of roughly 300 small cap companies. As of June 30, 2009, the Fund held roughly 60 of those names with annual turnover on the high side of approximately 160%. Click here to download the PDF.

Wall Street Journal’s print edition on October 5, 2009 also featured Mr. Cinnamond as placing second in the Winner’ Circle contest for the 12-month period through September 2009 (posting a 29% return).

We decided to take a closer look using daily data and returns-based style analysis (RBSA), and found Mr. Cinnamond’ selection skills over the past eighteen months have been strikingly strong relative to other small capitalization mutual funds.

We were surprised to find timing contribution was only incremental given Mr. Cinnamond’s sizeable moves into cash and other sectors over this period of high market volatility. Most of the excess performance as determined by the sector specific returns-based style analysis model was generated through selection as shown in the chart above. Together, the combination of both timing and selection skill led to dramatic excess outperformance of 23.0% over the Russell 2000 Index for the period March 2008 to September 2009.

Click here to download the PDF.

MPI clients: To obtain the mpi Stylus Pro daily attribution template, please contact support@markovprocesses.com.