The Curious Case of Bowen, Hanes and Tampa Police and Fire Retirement Plan: Part 3
In nearly 40 years Bowen Hanes has generated an annual equity return of 15%, versus 12.4% for the S&P 500, on behalf of the Tampa Police and Fire Retirement Plan. The firm has beaten the index in 28 out of 40 years. That’s an outstanding record. Like many of the portfolio managers I discussed yesterday, Bowen got off to an amazing start by beating the market 12 years in a row (1974-1986). According to my calculation, during those formative years Bowen beat the S&P 500 by about 5.5% each year. Thereafter, Bowen has beaten the market by an average of about 2% through November 2014. In my experience, these results look realistic rather than fraudulent. A superior firm starts with great performance which ebbs over time. As I’ll discuss in this post, that doesn’t mean that Bowen’s results versus the S&P 500 are totally accurate.
For starters, successful managers often have a great beginning track record. If the results hadn’t been great at the outset, they wouldn’t have attracted notice in the first place. As a result, there is a lot of survivor bias in the universe of money managers. Only the winners tend to survive. So Bowen’s great start is actually quite normal, although the 12-year run is extraordinary.
Nonetheless, Bowen’s early results may be overstated, especially against the S&P 500. Let me emphasize that I haven’t talked to anyone at Bowen. Thus it is possible that Bowen made all the necessary adjustments to ensure that its track record is precise. However, in my many decades of experience as a money manager, I’ve found that historic track records are prone to exaggeration due to a variety of factors. With the advent of computers and more precise performance standards, these errors have become less common. I’m not suggesting that Bowen committed fraud. Rather, I’m hypothesizing that Bowen, like many money managers in the 1970s-1980s, made mistakes in calculating performance that exaggerated their margin over the S&P 500. In seeking to acquire money managers for NationsBank, Legg Mason, and my clients, I found problems in many performance track records. Here’s my list of possible errors in Bowen’s early record:
1. Distorting Cash Flows: Bowen calculated the returns on a time-weighted basis (just like a mutual fund). This is a perfectly acceptable way to measure investment performance. However, if there are significant client contributions or withdrawals during a reporting period, those cash flows can distort the returns. Mutual funds get around this problem by calculating the return every day and then chaining (multiplying) the daily results. In the 1970s and early 1980s this wasn’t easy to accomplish. Electronic spreadsheets, let alone proper investment performance software, didn’t exist. In my view, there’s some likelihood that Bowen’s performance record could have been distorted a bit by contributions from the Tampa pension plan, which was expanding.
2. Allocation of Cash in Balanced Accounts: Since Bowen was managing a balanced account (stocks and bonds) with changing cash balances, the allocation of cash could have distorted Bowen’s returns. Back in the early days of money management, cash could be moved around to enhance investment returns. In a rising market available cash might be allocated to fixed income in order to boost reported equity returns, and in bad markets more cash might be shifted to the stock account in order to reduce the reported loss in the equity account. Again, we don’t know how Bowen treated cash. However, in my experience cash policy was arbitrary in the early days of money management and tended to exaggerate equity performance.
3. Somewhat Mismatched Benchmark: In the 1970s and early 1980s, the S&P 500 was the only benchmark form measuring equity performance, and style (growth and value) and market cap (large and small companies) weren’t measured. Thus, many of the successful money managers of the early years had great freedom to pick stocks that weren’t representative of the S&P 500. As I mentioned in my previous post, the consultant community began to rein in this freedom, and over time the managers selected securities that more closely resembled their benchmark. Bowen, Hanes has resisted this approach over the decades, which is evident in the variation of their correlation and tracking error versus the S&P 500.
While trying to verify Bowen, Hanes investment track record, Leanna Orr, Managing Editor of Chief Investment Officer magazine, asked Jay Bowen for a copy of the firm’s composite returns. Although Mr. Bowen claimed that the firm’s composite adheres to industry standards (Global Investment Performance Standards), he refused because the figures hadn’t been properly adjusted for fees. I think this is a lame excuse. I suspect that there’s another reason (again, it isn’t fraud).
Bowen, Hanes manages a lot of money for taxable investors, and probably customizes portfolios. This was the old school way of managing money before the consultants entered the picture. In addition, there’s a hint in Part 2 of the firm’s ADV filing with the SEC. Bowen, Hanes doesn’t block trade. They tend to buy securities for each of their 171 accounts. As a result, I’ll bet that Bowen, Hanes has multiple composites and a fair amount of dispersion within those composites. Ms. Orr reported that a couple of Bowen, Hanes Florida accounts reported differing results. I suspect they don’t want to air those details.
Bowen’s critics are using modern standards of investment management and disclosure for a firm that is still operating in the 1970s. By modern standards, Bowen’s performance record, investment practices, and disclosure standards aren’t appropriate, and therefore are suspicious. Those suspicions don’t, in my view, add up to fraud or a ponzi scheme, as critics imply.
Tomorrow we’ll conclude this series with a discussion of the shortcomings at Tampa’s pension and the role of consultants.