Investment Process

Identify the right companies

Our investment process starts by choosing the right companies.  The single most important decision we make is which of the companies among the thousands of public companies deserve our ongoing time and attention.  We strive to select good companies - sound businesses with strong management that drive their intrinsic value higher over time, even in difficult periods such as the credit crisis of 2007/2008.  During this crisis, many companies went bankrupt and many more were forced to issue equity at very low prices.  In contrast, we owned companies that had (and continue to have) strong balance sheets and that, in many circumstances, were able to take advantage of the financial crisis to build long term value, either through reinvesting in their businesses or by making strategic acquisitions.

Get to know the companies extremely well

After choosing the right companies on which to focus, we spend considerable time and effort formulating and maintaining a view of each company’s value.  Our background, skill-set, and approach - spending our time on the right things - differentiate us from other investment managers.  We enjoy the long and hard process of really understanding a company.  We believe that our discipline and the depth of our analysis is rare.

Build and maintain an optimal portfolio

Having spent the time to understand each of our company’s value, we then construct an optimal portfolio.  Simplistically, our optimal portfolio is diversified over approximately 25 companies in largely un-related industries with the size of each holding determined by many factors, the most important being how attractively priced it is.  Changes to our portfolio over time are primarily driven by our response to changes in traded prices - the cheaper a company gets the more of it we own, the more expensive a company gets, the less we own.


We believe risk is the prospect of permanently losing money.  This contrasts with most investors and academics who describe risk as the likelihood and magnitude of prices bouncing around.  As such, over the past few market cycles, there has been an increasing obsession with finding ‘uncorrelated assets’ in order to build a portfolio that exhibits as little measured price volatility as possible.  The problem with this approach is two-fold.  First, it presumes that shorter term price volatility equates to risk and, second, it presumes that historical relationships between different asset classes will hold in the future.  Time and again, the second presumption is refuted as historical relationships between assets go haywire (often at just the wrong time).  But more importantly, the first presumption is simply wrong.  Risk is not volatility; it is the prospect of permanently losing money, not a short term quoted price decline.

But before we go further, we need to make one important point:  discussing risk in the absence of a time horizon is meaningless.  An investor’s time horizon is absolutely essential to understanding his or her capacity to handle risk.  If you absolutely need to have access to all of your money instantaneously, then you have a very low time-risk tolerance and the prospect of a short term quoted price decline is your definition of risk.  As such, you should have all of your money in treasury bills.  If, like most investors, you have a longer term horizon, you should measure risk (i.e. the chance of losing money) as a decline in unit price, measured over a reasonable period of time.

So over what time period should declines in unit price be measured?  We believe three years is the minimum period over which to evaluate any investment manager since shorter periods can give very misleading results – a manager may be brilliantly building positions in undervalued companies or may be stubbornly pouring good money after bad in overvalued companies.  Three years is the point at which one starts to cross over from the ‘voting’ to the ‘weighing’ characteristics of the market; although five to ten years is required for the real shift to occur.

Since Turtle Creek’s founding in 1998, there have been 225 three year rolling periods (i.e., Oct 31/98 to Oct 31/01, Nov 30/98 to Nov 30/01, . . .).  Over these periods, market1 was negative 19% of the time.  We, on the other hand, were negative only 7% of the time.

Negative Three Year Returns

Market1: 43 of 225 (19%)
Turtle Creek2: 15 of 225 (7%)

And lest you think our negative periods were any more severe than those suffered by the market, in fact it's the opposite:  the average annual return over our 15 negative three year periods was -7.4% whereas the market's average annual return over its comparable worst 15 periods (worst 15 of a total 43 negative periods) was -8.5%.

So, by the numbers, one can see that we have been pretty successful in managing risk.  But just as importantly, this is where all of our money is and we are conservative investors.  We manage risk by understanding the value of the companies we invest in and then constructing a diversified portfolio that, among other things, owns relatively more of the most attractively priced companies.  Put simply, the cheaper something is, the less risky it is.

1. The market’s performance from November 1, 1998 until December 31, 2015, reflects the performance of the S&P/TSX Composite. From January 1, 2016 to December 31, 2018, the market’s performance reflects the return from a 75% weighting in the S&P/TSX Composite and a 25% weighting in the S&P MidCap 400. From January 1, 2019 onward the market’s performance reflects the return from a 50% weighting in the S&P/TSX Composite and 50% weighting in the S&P MidCap 400. Beginning on September 1, 2003, an annual fee of 10 basis points has been applied to the benchmark, S&P/TSX Composite and S&P MidCap 400 on a monthly basis, to reflect the approximate costs of investing in a security that aims to track an index or benchmark. The manager feels a blended benchmark, with varying weights, is appropriate because the weights noted above roughly correspond to the average country exposure of Turtle Creek during the same periods. Prior to December 31, 2015 Turtle Creek’s average U.S. company exposure was less than 3%. From December 31, 2015 to December 31, 2018 Turtle Creek’s average U.S. company exposure was 29% and since December 31, 2018 it has averaged 50%.

2.  Turtle Creek's performance is of Turtle Creek Investment Fund Class A Series 1 Units ("TCIF") to November 1, 2008 and Turtle Creek Equity Fund Class I Series 1.0 Units thereafter and is net of all fees, carried interest, and expenses.  Since Turtle Creek Equity Fund maintains an almost identical portfolio with that of TCIF (with the exception of TCIF’s private company investments), historical performance for TCIF has been combined with that of Turtle Creek Equity Fund.  There were no private investments in TCIF before 2003 and, in aggregate, the private investments had a minimal impact on TCIF's returns to November 1, 2008.  TCIF’s carried interest structure is not the same as the structure used for Turtle Creek Equity Fund (details available upon request).

All data is in a common currency (Canadian dollars).  S&P/TSX Composite and S&P MidCap 400 are total return indices.  Comparisons to certain indices are provided for illustrative purposes only, and are intended to indicate broad market performance.  Comparisons to indices are limited because indices are not managed and do not charge fees or expenses.  The Fund may underperform or outperform the indices for many reasons.  

Past performance must never be construed as investment advice or a prediction of future performance.

Source: Turtle Creek Asset Management, Bloomberg.  As at June 30, 2020.

Intrinsic Value

We spend the great majority of our time working to understand the value of each of our investments and avoid trying to guess what each stock price will do.  While many managers speak of intrinsic value, our observations are that precious few actually take the time to really think about it; instead, relying upon relative valuations to provide them with a view of value.  Of course, the problem with comparing the prices of a company with its competitors is that it doesn't really help you if a sector is either overvalued or undervalued.  Just ask technology investors from the late 90s how effective comparable analysis worked for them.

So how do we arrive at our estimates of intrinsic value?  We start by assessing the risks and opportunities available to each company and then incorporate those assessments into a financial model.  We think about the factors that will determine a company's future cashflows:  its revenue growth, margin opportunities and threats, operating expense requirements, tax rates, capital expenditures and working capital requirements to name a few.  We look at returns on capital and operating margins as sanity checks.  Using this model, we discount the company’s future cashflows to arrive at a current value.  This value roughly coincides with the price a buyer would pay for the entire entity.  The real value of doing all the work to build a model is that it consolidates a great wealth of historical information and forces us to think long and hard about all the factors that will affect the future of the business.  It forces us to explicitly consider all crucial assumptions and is a guard against untested speculation and optimism.

While developing a view of intrinsic value requires a great deal of effort, it prepares us for the events that inevitably occur (either market related or specific to the company) that cause meaningful changes in share prices.  The grounding that we have from our extensive work to arrive at a view of fundamental value allows us to react during these times.  Buying more of a company when its share price has fallen sharply isn’t that difficult to do if you truly understand the company’s value.  The truth is that most investors do not understand value and so are not comfortable ‘stepping-up’ at the times they should.

When to Invest?

We do not try to time the market and we think investors shouldn’t attempt this either.  Furthermore, we do not take a stance on whether or not the stock market is overvalued or undervalued.  Instead, we take a stance on whether or not each of our companies is overvalued or undervalued.  There have been a few periods of high valuation where we were not fully invested simply because the price of some of our companies rose to such a level that we were no longer interested in owning very much.  In particular, during the buoyant markets of 1999 and early 2000 we had net cash of over 30%.  But this was an anomaly:  it is an unusual situation where we cannot find a quality company to own that offers a superior risk return proposition relative to investing in treasury bills.  To be sure, if we could time the market and knew for certain that stock prices were going to fall, then we would simply sell everything, wait for the decline and then reinvest all of our money.  We do not believe, however, that it is possible to forecast the direction of the stock market in the short term.  Therefore, as long as we can find companies that are generating strong cash flows, have excellent growth prospects and are trading well below intrinsic value, we are likely to be a fully invested fund.  We believe that the best way to earn superior investment returns over the long run is to stay invested.