Tuesday, July 23, 2019

Keweenaw Land Association

Keweenaw Land Association


Keweenaw owns 184,000 acres of timberland and 401,841 acres of mineral rights, mostly located in the Upper Peninsula of Michigan. This type of investment is what I would describe as an asset play, the company’s value is really in the difference between the value of the timber it owns and the price you can buy the stock for. The market is pricing the company at a significant discount to the value of its standing timber, despite timber being an asset class that, long-term, delivers returns similar to those of the S&P500. 

There’s 1.3M shares outstanding and 184,000 timberland acres, that’s 0.14 acres per share. The company has an appraisal done every 3 years of the value of its timberland.  In 2018 an appraisal was done and depending whether you use a DCF or a sales comparison approach you get different values. Using the sales comparison approach – which uses recent transactions as a guide – the timberland is worth $160M. Or $870 per acre. Multiplied by 0.14 you get $122 a share. Subtracting net debt, you get $112 a share of timber. The market currently values the company at $71 a share. 

The reason I prefer the sales comparison approach is because it’s based on sales that were done in the past in the very region Keweenaw owns land. The income capitalization approach, which is essentially a DCF, relies on the appraiser making much more subjective assumptions about discount rates, etc. Different discount rates will make the value of the timber fluctuate very widely. 

Timber is a unique asset because, over time, it will appreciate in value due to the growth of the tree along with the fact that the underlying land also increases in value. Whilst prices can be volatile, returns for owners don’t need to be. The harvest is like the dividend yield on a stock. Owners can simply reduce harvest and “save” the earnings for later, similarly to how a company can retain earnings and put them to work at a more opportune time. 


Keweenaw’s background traces back to the period after the Civil War, when Congress provided a land grant to help finance the construction of a shipping canal across the Keweenaw Peninsula of Upper Michigan. 

A company by the name Portage Lake & Lake Superior Ship-Canal Company was supposed to build the canal. But after experiencing financial setbacks, the assets of the company, including the 400,000-acre land grant, were purchased in a bankruptcy auction, by the company's creditors. From there, the Lake Superior Ship Canal Railway and Iron Company was formed. In 1891, that very company sold the completed ship canal to the U.S. government, and the remainder of the assets, including the land grant, was transferred to the company’s successor, the Keweenaw Association, Ltd. Later the company was renamed Keweenaw Land Association.

Timber as an Asset Class

Timber is usually classified into either softwood or hardwood. The name, however, doesn’t have anything to do with the hardness of the timber, it merely indicates which species of tree the timber comes from. Around half of US softwood timber goes into new home construction, and one third goes into repair/remodeling market. In terms of US hardwood, which covers most of Keweenaw’s land, around half is used for furniture, cabinets and flooring and molding/millwork. The company’s standing timber inventory in terms of volume (in cords) is as follows: 81% pulpwood and 19% sawtimber. Even though pulpwood makes up the bulk of the volume, sawtimber makes up the bulk of the value.

Timber returns haven’t been very good in the last few years compared to the period before 2007. From 1987 to 2007, returns were in the order of 10% p.a., whereas in the period between 2008-2015 returns were ~4% p.a. What contributed to a decline in returns? One factor was the decline in housing construction, which lowered high-quality timber prices. In the U.S. housing starts went from 2M in 2005 to about a quarter of that in 2009. The second factor was really low inflation since the recession. In half of the years when timber had high returns inflation was over 3%. For comparison purposes, over that same period the S&P500 returned 12% p.a.

Return in timberland = (Profit from harvest / appraisal value) + tree growth rate after harvest + rate of inflation.

In 2018 more timber was harvested than in any other year. There's a question of whether management will cut trees down too fast or not. If the trees grow 4% a year, then they should have grown by 160,000 cords in 2018. Management harvested 70% of the growth. So, the trees grew 1.2% after harvest, the (profit / appraisal) adds another 1.8% return, and inflation was 2%. All in all, the return was 5%. With an inflation rate of 2-3% returns can range from 5% to 7%.


Keweenaw sells different kinds of timber, and the price between the most expensive (which accounts for 2% of sales volume) and the least expensive is wide. Over three-quarters of KEWL’s sales volume is pulpwood, (i.e. used to make paper and cardboard boxes), which is the cheapest kind of timber. The most expensive kind of timber might be used for furniture, flooring, etc., and this type of timber may be exported. On the contrary, cheaper types of timber will not be exported or taken by trucks to far away mills because the cost of transportation is very significant. So, like with a cement company – despite cement being a commodity like timber – there is a locality-based moat, because of the cost of transportation. 

Given the implications above, almost 2/3 of the volume of their timber is sold to regional mills. Contracts with mills range from 3 months to 2 years, which set a specified number of logs and pulpwood to be sold with the price is subject to market pricing at the time. 

Overall, in 2018 the company sold over 111,000 cords (1 cord is 3.6m²), which resulted in more than $14.2M in sales. The cost of producing this was $10.4M, less SG&A which was about $1.6M, you get an operating income of $2.2M. EBITDA from timber operations was $2.7M. Both operating and gross margins have been steady. Since 2003, the company only lost money in 2003, because of over $2M in non-recurring expenses associated with the Cornwall Capital proxy fight, settling employee contracts and costs associated with investigating an eventual conversion into a REIT with an investment bank. 

The company’s revenue averaged almost $10M between 2011 and 2015, and over $12M between 2016 and 2018. Gross margins are 30%+ and operating margins are 10%+. In recent years both revenue and operating margins have expanded and that’s due to increased harvesting. Harvesting from the company’s lands decreased very substantially, under previous management, when prices for timber softened due (1) drop in housing starts and (2) low inflation. If we go back to 2011 the harvest volume was almost 90 thousand cords and it went down to 70 thousand. Since then it has gone up to 110 thousand cords, with a substantial increase last year when Cornwall Capital took control of the company. 

What is the whole company worth?

The timberland is worth $160M and there’s $5.6M in cash and $18M in debt. This doesn’t include mineral royalties associated with the Highland Copper Company, where initial production is expected in 2021. In total you have a company that’s worth $112 share, and you can get it today for $71 a share.

The company is cheap, now the question is whether they can service the debt. The company has a lot of debt relative to its cash earnings, but not by the lender’s standards, who measures indebtedness as a percentage of the timber value.

The Debt

Keweenaw’s debt breaks down as follows: $13.7M out of $25M drawn in a five-year revolver, due on December 2021. This LoC was used by previous management to buy land in 2017. Because more than half of the total line of credit has been drawn, it carries interest of 4% p.a. The portion that has not been drawn carries an interest of 0.0875%. This line of credit carries a covenant whereby the principal balance cannot exceed one third of the value of the timberlands owned by Keweenaw. Using the $870 per acre value of the timberland, the principal balance is less than 10% of the value of the timberlands. 

There's also an option to term out the loan. Essentially the repayment period for the long can be extended. Whilst I don’t know into how long the loan would be converted into, I know that if it was converted today, the company would pay 3.5% p.a. in interest on the $13.7M.

In addition to this LoC, the company also has a $4.5M loan, due on December 2026, with carries a fix interest rate of 3.05% p.a. Including this $5M loan, debt to timberland value, is less than 12%.

In the last 7 years, the lowest EBITDA the company generated from timber operations was $1.2M and the highest was $2.7M. The highest being in 2018 was due to higher volume, and in 2015 it was due to higher high-quality timber prices. The highest was $2.1M in 2015, and whilst volume was almost a third lower than it is today, prices were higher, and the increase prices translated into a larger gross margin that fell to the bottom line. The EBITDA margin has ranged from 13% to 24% over those same 7 years.

If prices remain constant and harvesting decreases to 80,000 cords (almost 30% less than 2018), EBITDA would be north of $1M. This seems reasonable given that I don't think Cornwall will be under-harvesting.

Investors might worry about the debt and claim that because of it, the company deserves to trade for less than its assets are worth. If we assume there’s a 50/50 chance of the company going bankrupt, then we should apply a 25% discount to net asset value. Which would mean the stock would be worth around $84 a share. But I think the odds the company has of surviving are more like 90/10, which means of 9% discount is appropriate. 

The Key Variables

The first key variable is servicing debt. If debt isn’t serviced none of the other variables matter. Debt has been discussed above. The second and third key variables are (1) inflation and (2) housing starts, because these inputs will affect higher-quality timber prices – of which Keweenaw has a lot – over the next few years. High-quality timber accounts for 23% of volume and almost 40% of sales. The more profitable timber is the one most related to housing construction, so housing becomes and important input for prices.

Other variables include the fact that prospective buyers of timberland have been net sellers for the past few years. And so, management thinks this is not the right time to sell the company. For the past 18 years sales to REITs or Institutional investors have ranged from 300,000 to 600,000 acres in size, so there is a market for KEWLs large piece of land to be sold in one go. There’s also been a notable trend over since at least 1995, in which timber land has shifted from paper/lumber companies to investors/REITs. However, since 2006 transactions have been of smaller nature, with 2 notable exceptions in which in each case more than 280,000 acres were sold. Today, investors are looking for high-quality properties which fares well with KEWL’s high-value northern hardwood.

Lastly, there’s the risk of Cornwall Capital selling out a price below what the company is really worth. Cornwall manages $320M (according to SEC's Investment Advisor Public Disclosure) has a 26% stake in the company, and as of 2018 they control the company. This may be a catalyst. However, there’s also no way Cornwall can sell their position over a reasonable period of time considering the dollar volume of shares transacted each day. Because of that, they may get “rid” of Keweenaw at a lower price than a owner would accept. 


The first question on my mind is whether Keweenaw will be able to service its debt. If interest expense runs at $700k per year, there’s enough cash to cover 8 years of interest expenses.  Besides that, EBITDA from timber operations was the lowest in 2013, at $1.2M, when prices were roughly 10% lower (production mix was very similar) and harvesting was less than two-thirds of what it is today. So, I think it’s reasonable to assume EBITDA of more than $1M. 

Cornwall has also showed an appetite towards monetizing non-core assets (i.e. the land the company owns not used for timber). In 2018 land sales were almost $1.5M. That’s as much as the previous management sold, combined, in the previous 7 years. There’s about 11,000 of non-productive land that will be sold to improve the balance sheet. I estimate that land, which is mostly conservation land and some lake/river frontage, is worth another $8M.

In terms of management I'm inclined to think it’s of higher quality than it was before Cornwall took control, but it's hard to make judgements based on a one-year performance.. Nevertheless, the previous directors that had been with the company for 25+ years didn’t do much for shareholders, except for entrenching themselves on the board. Board expenses were 2-4% of revenue. With all directors owning virtually no equity in the business, shareholders have been treated unfairly with a stock that hasn’t not appreciated in value for over one decade.

Timber is an attractive asset class that has had similar returns to the S&P500 over the long-term. By buying Keweenaw you’re buying timber at a large discount to its appraisal value. A stock should only trade at a large discount to NAV if, (1) the stock’s returns are going to be below the market’s long-term return or/and, (2) there’s a high possibility that most of the company’s value won’t flow to shareholders. 

Today you can buy Keweenaw for the same price as 12 years ago, when the appraised value per acre was $400. The timber, net of debt, is worth $112 a share, whereas the current stock price values it at $68 a share, ~60% lower. If the common stock is worth $112 a share, and you buy it at $68 a share, and after 3 years Mr. Market realizes it’s actually worth $112 a share, you’re looking at a 16% p.a. return.

I believe this type of a risk-reward situations is quite attractive. Keep in mind we’re using timber valuations per acre from 2018, which are the lowest in a long time. In 2012 there was three-quarters of today’s acreage and the valuation per acre 5% lower. In 2015 there was 10% less land and the valuation per acre was almost 10% higher than it is today.


  1. The company has released it's Q2 2019 report.

    Debt was reduced by $4.5M (from $18.5M to $14M), through a combination of cash (had $2.7M on hand), sales of land (sold $689k) and sale of marketable securities ($2.3M sold).

    The company was left with +$1.1M in cash (including sale of land in August).

    Updated Appraisal (basically unchanged):

    - Timberland $160M
    - Cash $1.1M
    - Debt $14.5M

    Total = $146.6M
    Per Share = $112

    (doesn't include land which is probably worth north of $7M)

    Because of the reduction in debt, interest expense has come down from my estimate of $700k a year to about $530k or so.