Timberland Investing: What You Need To Know Now

| About: iShares S&P (WOOD)

I. Executive Summary

This report provides a comprehensive and up-to-date look at U.S. timberland investing. It addresses:

  • Historical timberland returns
  • Expectations for the future
  • The impact of inflation on timberland returns

The following research was conducted by Forest Investment Associates (FIA), a timberland investment firm with $3.9 billion of U.S. timberland assets under management as of the 4th quarter 2012. Our conclusions show:

  • Timberland returns have varied over time as the industry has matured. Returns were negative for a year after the financial crisis of 2008, but have since been increasing.
  • Long-term demand drivers for timber remain strong and most indicators predict a meaningful mid- to long-term recovery in timber prices. This recovery should support stronger timberland investment returns in the future.
  • In the face of increasing demand, a significant reduction of timber supply is projected, primarily due to a pine beetle epidemic in western Canada.
  • Timberland returns show significant positive correlation with inflation, while stocks and bonds do not. Timberland returns have historically exceeded the long-term rate of inflation, while stock and bonds have not.

II. Historical timberland returns

Below is a chart of The National Council of Real Estate Investment Fiduciaries (NCREIF) Timberland Return index, which has measured timberland returns since the late 1980s.

Figure A: NCREIF timberland returns (1987 - 2012)

As shown above, timberland returns consist of current cash flow (represented by EBITDDA in the NCREIF index, shown in red) and appreciation (shown in green). Much of these returns have been appraisal based. We believe historical returns could be summarized as follows:

  • The early years (from the mid-1980s through the late 1990s) - Returns routinely exceeded 10% or even 20% annually, including a material portion of returns reflected in cash yields. Investor capital was quite limited, however, and the timberland investment industry was in its infancy.
  • Late 1990s through early 2000s (the "Rise of the TIMO") - Returns suffered as timber prices retreated in the recession of 2001, though capital flows into the asset began to increase and larger scale institutional timberland investment management companies ("TIMOs") were formed.
  • Early 2000s through 2008 (the era of the "Mega-Deal") - Returns increased substantially, but in large part due to appraisals on ever-higher prices for properties. Transactions grew increasingly large (multi-billion-dollar), as integrated forest products companies took advantage of new investor capital to deconsolidate their raw material supply chains.
  • 2008 to present ("After the Crash") - Over the last several years, returns have been disappointing, dropping off significantly after the global financial crisis of 2008. While cash flows have remained positive, asset values have eroded, which roughly tracks with the decline in timber prices (discussed below) in connection with the crash of the housing market in the U.S. As timber prices have begun to increase, asset values have begun to recover as well.

III. Timber prices

Timber prices over the last 25+ years have been characterized by a significant run-up through the early 1990s, followed by a flat to slightly downward trend subject to significant volatility. (See Figure B below for a chart of the primary major timber products in the two major timber producing regions of the U.S.: the Southeast and the Pacific Northwest.) The initial run-up in prices was driven, in part, by the removal of Federal timberlands from active harvesting (following the protection of the Spotted Owl). This action drove much of the superlative returns (both cash flow and appreciation) to early investors.

Figure B: Timber prices (1985 - 2012 YTD)

Source: Timber Mart-South, Log Lines, Oregon Department of Forestry, Pacific Rim Wood Market Report

Looking at these price trends over the more recent past shows that timber prices began to soften in the middle part of the last decade, followed by a precipitous drop around the financial crisis of 2008. (See Figure C below.) Over the last couple of years, Pacific Northwest timber prices recovered meaningfully. Southern prices, meanwhile, remained depressed. We believe that the primary drivers of the Pacific Northwest price recovery over the last three years have been (a) increased exports, particularly to China, and (b) additional regional demand pressures due to the pine beetle epidemic in Canada (discussed further below). These demand drivers have not been felt as significantly in the U.S. South.

Figure C: Timber prices (2003 - 2012 YTD)

Source: Timber Mart-South, Log Lines, Oregon Department of Forestry, Pacific Rim Wood Market Report

IV. Timberland valuations

Average per-acre timberland values are shown in Figures D and E below[1]. Over the long term, timberland asset values have increased, partly due to inflation and partly to a decline in discount rates over time as the asset class and the industry have matured. Over the recent past, asset values ran up just prior to the financial crisis, and softened thereafter. The relatively greater increase in timberland values in the Pacific Northwest is due to the same dynamics that have caused timber prices to increase in the region, namely, increased exports, particularly to China, and regional supply reductions associated with the Canadian pine beetle epidemic (discussed below).

Figure D: Regional timberland values (1987- 2012 YTD)

Source: Forest Investment Associates research

Figure E: Regional timberland values (2005- 2012 YTD)

Source: Forest Investment Associates research

V. Transaction volumes

Timberland transaction volume over the last 15 years (Figure F) appears to have roughly tracked timberland returns (Figure A), building during the "Rise of the TIMOs" in the late 1990s and early 2000s, increasing substantially through the mid-2000s in the era of the Mega-deal, and dropping off after the financial crisis. It is important to note that the era of the Mega-deal was made possible by a paradigm shift in which timberlands -- historically held by a relatively large number of vertically integrated forest products companies -- were sold to a number of newly formed TIMOs, which were able to fund the acquisition of these assets with equity capital from institutional investors who flooded into the market at that time.

Figure F: Timberland transaction volumes (1995 - 2012)

Source: Forest Investment Associates research

VI. Increasing demand

Timber demand and timber prices are driven, in large part, by demand from housing construction. Construction demand for softwood sawtimber represents approximately 65% of total demand for the product. Of this, about two-thirds relates to new housing construction, with the remainder reflecting demand associated with repair and remodeling. The chart below depicts U.S. housing starts over the past 50 years.

Figure G: Housing starts (1959 - 2012)

Source: U.S. Department of Commerce

Housing starts have never been this low. Historically, even in periods when housing starts declined dramatically, they recovered in less than a year. In contrast, housing starts have now been at record low levels for nearly four years. Figure H shows housing starts over the more recent past.

Figure H: Housing starts (2001 - 2012)

Source: U.S. Department of Commerce

The slowdown in housing starts not only reflects lower current demand, but presumably an over-supply of housing built over the last 10 to 15 years. The next chart (Figure I) shows annualized housing starts, filling in areas to illustrate the extent to which housing starts were higher or lower than the long-term annual demand of approximately 1.5 million housing units. Reconciling the areas above and below the 1.5 million starts line shows cumulative underbuilding or overbuilding relative to long-term trend.

Figure I: Quantifying overbuilding/underbuilding (1998 - 2012)

Source: Forest Investment Associates research

As shown above, for the period from the start to the end of the housing bubble (from 1998 to 2006), homes were overbuilt by a cumulative 2.3 million. Yet the housing collapse has now continued for so long, and been so deep, that on a cumulative basis relative to the long-term demand of approximately 1.5 million units, the level of underbuilding -- nearly 4.3 million new homes since 2007 -- is nearly double the amount of overbuilding that occurred during the housing bubble.

By either measure (the recent past or long term), the significant depression of new home construction of the last several years has counterbalanced any prior overbuilding. While nearly all economists forecast an inevitable housing recovery (discussed below), it will take some time to move from our current rate of housing starts to the long-term average, which will only serve to further increase the amount of net underbuilding.

Simultaneous with the collapse in housing starts, the share of Americans living in multi-generational family households is the highest it has been since the 1950s, having increased significantly in the past five years. In 2010, 21.6% of adults between 25 and 34 lived in this type of household, compared to 15.8% in 2000 and 11.0% in 1980 [2]. This "negative household formation" results in a near-term reduction of housing demand, but is not representative of the expected long-term trend.

As stated in a recent report by the Joint Center for Housing Studies of Harvard University:

…over the longer run, the most important drivers of household growth are the size and age structure of the adult population. Assuming the economic recovery is sustained in the next few years, the growth and aging of the current population alone - including the entrance of the echo boomers into adulthood - should support the addition of about 1.0 million new households per year over the next decade. The biggest unknown is the contribution of immigration to overall population growth. But even assuming net inflows are roughly half the level in the Census Bureau's 2008 projection, the Joint Center for Housing Studies projects household growth should still average 1.18 million a year…[3].

Using this 1.18 million figure as a base, and accounting for both net removals as well as vacant unit demand, housing starts are expected to average a minimum of 1.65 million units, as Figure J illustrates.

Figure J: Harvard projected new home demand (2010 - 2020)

Source: JCHS Working Paper W10-9. Low immigration is half Census Bureau baseline immigration assumption. High immigration is Census Bureau baseline immigration assumption.

Over the near term, what might a housing recovery look like? Figure K below shows a composite of housing starts projections through 2015.

For the timberland investor, it is important to understand that asset values -- and returns -- are predicated on projected cash flows from harvests over a period of several decades or more. The fundamental long- term housing demand as illustrated in Figure J is therefore more relevant than the specific timing of a housing recovery, as shown in Figure K below.

Figure K: Housing start forecasts

Source: various

The implications of consistent and strong long-term demand (as shown in Figure J) and the current and near-term projected pace of the housing recovery is a very significant amount of pent-up demand. While we do not believe housing starts will reach a sustained level above 2 million, a simple mathematical exercise in which we reconcile the significant under-building in the first half of this decade with the 2010-2020 cumulative demand for housing shown in Figure J above suggests that when pent-up demand is released, the strength and magnitude of the housing recovery will be significant.

Figure L: Theoretical housing starts in the second half of this decade

Source: Forest Investment Associates research

VII. Reduced timber supply

Coinciding with a significant projected recovery of demand, an unprecedented reduction in timber supply is occurring in Canada as a mountain pine beetle infestation has reduced timber volumes by more than half in the western half of the country. This ecological disaster will have a material effect on the United States.

Figure M: The Canadian Pine Beetle epidemic

Source: Natural Resources Canada

The pine beetle is expected to kill up to one billion cubic meters of timber in Canada's British Columbia, the heart of the epidemic. Canada provides the U.S. with one-third of its lumber needs, with the majority of that coming from western Canada. Thus, the widespread loss of timber to the Canadian Mountain Pine Beetle will result in appreciable extra demand -- and strain -- on U.S. timberlands.

One way to evaluate the scale of the reduced supply is to equate it to the scale of demand -- in terms of number of housing starts. Figure N below illustrates that the reduction in annual supply from Canada would equate to the same amount of wood that would be consumed in approximately 400,000 housing starts. Further, exports to China -- which effectively reduce supply that would otherwise be available for U.S. based demand -- equate to approximately the demand from 200,000 housing starts.

Figure N: Supply factors equated to housing start demand equivalents

Source: Forest Economic Advisor, Plum Creeks

What, then, is the net impact of the supply and demand factors discussed above? Future timber prices are expected to rise substantially from their current historic lows. The following chart reflects forecasts for Southern Yellow Pine and West Coast Douglas-Fir over the next 15 years from Forest Economic Advisors (FEA), the dominant global forest products economist and research firm providing long-term price projections to the industry.

Figure O: FEA's long-term timber price forecast (through 2027)

Source: FEA

As this chart suggests, better days are ahead for timber prices. The more bullish projection for the West Coast Douglas-Fir compared to the Southern Yellow Pine is largely a function of improving regional lumber demand due to the pine beetle-driven collapse of the industry in Western Canada. This should allow Pacific Northwest producers to capture a larger share of markets versus producers in the U.S. South.

Timberland's inflation hedge characteristics

Historically, timberland, like other real assets, has been a consistent inflation hedge. The following shows the extent of timberland's hedging characteristic.

The first analysis shows the correlation between timberland returns [4] and the Consumer Price Index (CPI) in 10-year periods (the presumed average life of a timberland investment), beginning with an investment made in 1960 and exited in 1969. In that time frame, correlation was nearly perfect. The second bar on the chart below shows the correlation over an investment horizon beginning in 1961 and ending in 1970, and so on.

Figure P: 10-year rolling correlation between timberland returns and CPI (1960 - 2012)

Source: Forest Investment Associates research

This chart confirms that the correlations are positive over nearly every time period. For the sake of comparison, the next chart shows the correlation between CPI and financial assets, both equities and fixed income, over 10-year investment horizons.

Figure Q: 10-year rolling correlation between financial assets and CPI (1960 - 2012)

Source: Forest Investment Associates research

This chart shows that, with the exception of returns over the last couple of trailing 10-year periods, the correlation of financial assets and CPI has been consistently negative.

The second analysis looks at timberland returns from a different perspective. It compares compound annualized returns for each 10-year investment period with the annualized inflation rate for the corresponding period.

Figure R: Compound annualized timber returns and inflation rates (1960 - 2012)

Source: Forest Investment Associates research

As shown above, compound timberland returns have always exceeded inflation. In no instance on the above chart is a green bar (representing the annualized timberland return) lower than a black and white bar (representing the annualized inflation rate).

The same analysis looking at equities (here, the S&P 500 index) yields different results. Figure S below shows that annualized returns on equities were often lower than the annualized rate of inflation over the same period.

Figure S: Compound annualized equity returns and inflation rates (1960 - 2012)

Source: Forest Investment Associates research

While equity investors have almost always made money over a 10-year horizon, those returns have not always beat inflation. In the 1970s, for example, while 10-year returns were positive, they lagged substantially behind inflation.

Analyses of long-term government bonds vs. inflation yield a similar result, as seen in Figure T.

Figure T: Compound annualized government bond returns and inflation rates (1960 - 2012)

Source: Forest Investment Associates research

As this chart shows, bond investors before the 1980s lost money to inflation.

VIII. Summary

While timberland returns have been weak the last few years as housing starts have fallen to historic lows, supply and demand factors remain extremely favorable for the asset class. Housing starts, having been at depressed levels for such a long time, are poised for a very sharp recovery, and may even reach unprecedented levels (above two million starts annually) for several years in the second half of this decade. At the same time, on the supply side, the Canadian Mountain Pine Beetle has reduced available supply coming from north of the border, which historically has accounted for approximately one-third of the U.S. lumber market. Add to that dynamic the steadily increasing demand from China, and all signs point to a significant improvement in timber prices over the next several years. This is expected to not only drive premium returns for investors, but continue to provide them. Our research underscores Forest Economic Advisor's (FEA)'s bullish timber price forecast over the long term. Further, timber looks to be a very effective hedge against inflation.

[1] Source: NCREIF Timberland Index

[2] Pew Research

[3] The Joint Center for Housing Studies of Harvard University: The State of the Nation's Housing 2012

[4] Timberland returns as reflected in the NCREIF timberland index (from Q4 1986 to present) and the Wilson timberland index (from 1960 through Q4 1986)

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

Business relationship disclosure: Partner of Forest Investment Associates Timber Partners (FIATP)