Historical market and landuse dynamics is the SE US: Implications for actual and projected futures. 09/27/2016 GHG Sherphardstown 1

Transcription

1 Historical market and landuse dynamics is the SE US: Implications for actual and projected futures Bob Abt Karen Abt 09/27/2016 GHG Sherphardstown 1

2 Starting Conditions Demand (Recession) By Product Housing Paper Pellets Supply (Peak Plantations?) By Product Plantation Age Class Distribution Prices Product Price Ratios Projections Rent Implications Forest Carbon Implications Forest vs Ag Landuse Plantations vs Natural Rotation Length 09/27/2016 GHG Sherphardstown 2

3 Wood Products User-defined, dynamic, and linked Small Trees pulpwood, 5-9 diameter, years old, $6-$10/ton stumpage prices (now at top end of range) fiber, energy pellets, pulp Medium Trees chip n saw, 9-11 diameter, years old, $10-$20/ ton (now at top end of range) fiber and small sawn wood Large Trees sawtimber, 11+ diameter, 20+ years old, $25- $40/ton (now near bottom end of range $25-$30/ton) lumber, plywood Logging Residues 20% of pine harvest, 40% of hardwood harvest residues are in the eye of the beholder (or logger) 09/27/2016 GHG Sherphardstown 3

4 SUPPLY (AGE CLASS DISTRIBUTION AND LAND BY FOREST TYPE) 09/27/2016 GHG Sherphardstown 4

5 How the current state of the inventory drives the supply side State of the current forest inventory Forest type mix and age class distribution For plantations Age Structure: Defines supply for the next years Planting today driven by harvest opportunities for trees planted years ago Planting today drives pulpwood availability years from now And sawtimber availability years from now And markets for sawtimber at that time drive planting for the next wave... 09/27/2016 GHG Sherphardstown 5

6 Current environmental NGO interest is here Profit max assumption might be OK here 90% Privately Owned 95% of harvest from private land 2/3 of pine harvest from plantations 2/3 hardwood or mixed hardwoods 09/27/2016 GHG Sherphardstown 6

7 Forest Type and Age Class Distribution of Acres and Harvest 09/27/2016 GHG Sherphardstown 7

8 Acreage of Tree Planting Tree Planting in the South Southern Tree Planting, All States and Ownerships, ,000,000 2,750,000 2,500,000 2,250,000 2,000,000 1,750,000 1,500,000 1,250,000 1,000, , , ,000 0 Virginia Texas Tennessee South Carolina Oklahoma North Carolina Mississippi Louisiana Kentucky Georgia Florida Arkansas Alabama Fiscal Year 09/27/2016 GHG Sherphardstown 8 Source: USFS, GFC, TMS

9 Latest 30A plantation acres State Mil Planted Acres GA 7.17 AL 6.99 MS 5.41 FL /27/2016 GHG Sherphardstown 9

10 Current Plantation Age Class Structure State Totals Mask Big Local (Timbershed) Differences Growth/Removal Implications 09/27/2016 GHG Sherphardstown 10

11 Forestland trend stable, but not static >5% Forest to Ag >5% Ag to Forest Financial returns to landowners empirically significant explanatory variable 09/27/2016 GHG Sherphardstown 11

12 Total Timberland Stable but Not Static 09/27/2016 GHG Sherphardstown 12

13 What causes this? we identified the rise in timber net returns as the most important factor driving the increase in forest areas between 1982 and This is consistent with reports that the increase in forests largely involved timberland acreage. (Lubowski et al. 2008) This is a privately owned landscape where marginal agriculture competes with forest land both at the intensive (plantations) and the extensive (fallow agriculture) margins. 09/27/2016 GHG Sherphardstown 13

14 DEMAND 09/27/2016 GHG Sherphardstown 14

15 US SE Pulpmill Inputs /27/2016 GHG Sherphardstown 15

16 Softwood Lumber Production SE production up from 51% of 2006 peak in 2009 to 87% of peak in /27/2016 GHG Sherphardstown 16

17 Is Canadian Capacity Shifting to US SE? 09/27/2016 GHG Sherphardstown 17

18 TIMBER PRICES 09/27/2016 GHG Sherphardstown 18

19 Southwide Prices timber-mart south Pine sawtimber stumpage markets have not recovered Hardwood markets stronger than pine 09/27/2016 GHG Sherphardstown 19

20 Southwide Prices timber-mart south PST/PPW ratio which drives management strategy (thinning and final harvest) declines from 5:1 to 2.5:1 09/27/2016 GHG Sherphardstown 20

21 Price changes lead to Changes in forest land rents, which leads to Less plantations? Less timberland? Shorter rotations? Pulpwood as a key rent driver? (pellets) 09/27/2016 GHG Sherphardstown 21

22 INITIAL CONDITIONS - SE GEORGIA 09/27/2016 GHG Sherphardstown 22

23 09/27/2016 GHG Sherphardstown 23

24 SE-GA Timberland Area Trends 09/27/2016 GHG Sherphardstown 24

25 SE-GA Demand/Consumption Dynamics Every tree can feed multiple markets Change in one market affects all markets Example: pine sawmill residues and pine roundwood consumption Recession makes the current starting point atypical; this matters for thinking about baselines. 09/27/2016 GHG Sherphardstown 25

26 PST/PPW Ratio $/ton SE GA Stumpage Prices SE GA Pine Stumpage Prices $50 $40 Pine Sawtimber Pine ChipNSaw Pine Pulpwood $30 $20 $10 $- Source: Forest2Market 10 SE GA Pine Sawtimber (PST)/Pine Pulpwood (PPW) Price Ratio Source: Forest2Market /27/2016 GHG Sherphardstown 26

27 INFLUENCE OF MARKETS ON SE GA PROJECTIONS 09/27/2016 GHG Sherphardstown 27

28 3. Cohort (age class tracking) modeling of age classes with current plot based growth/removals 09/27/2016 GHG Sherphardstown 28

29 Using Expected Removals Rather than Current removals (recession recovery) 09/27/2016 GHG Sherphardstown 29

30 4. 3 with expected removals no economic adjustment 09/27/2016 GHG Sherphardstown 30

31 5. As above with economic adjustment but no land use change Economic adjustment means that high prices lower demand, may shift demand outside of region (leakage) or take demand away from other users (displacement). Markets respond to scarcity. Removals don t go up much 09/27/2016 GHG Sherphardstown 31

32 6. 5 with economic adjustment and land use change Land use change means timberland loss/gain is affected by timber prices 20 year long wave cycles start to emerge 09/27/2016 GHG Sherphardstown 32

33 How does Pine Sawtimber Inventory Affect Stumpage Prices? 09/27/2016 GHG Sherphardstown 33

34 What we can say about pine sawtimber (big trees)? In may areas of the South growth exceeds removals because: 1990 s planting boom is now sawtimber Recession in housing means prices down 40%, production/harvest down 30% (production recovering, prices not so much) Pine sawtimber inventories are expanding quickly projected 20% higher southwide 09/27/2016 GHG Sherphardstown 34

35 What we can say about pine sawtimber (big trees)? Why is this important: Landowners are postponing sawtimber harvest until prices recover Build-up in inventories will dampen prices when demand recovers Low pine sawtimber prices are the best predictor of how forest and agriculture compete for land in the South. Low pine prices/high agriculture prices mean less timberland. Postponed harvest means drop-off in planting will continue 09/27/2016 GHG Sherphardstown 35

36 What we can say about pine pulpwood (small trees)? In many areas of the South there are fewer trees in the pipeline Planting has dropped 40% since 2000 Pine plantation acres have stabilized Since industry restructuring (land separate from mills) management intensity is more price responsive Postponed sawtimber harvest means drop-off in planting will continue Current PPW prices too low to keep land in forest 09/27/2016 GHG Sherphardstown 36

37 Structural Change? Historical relationships (built into empirical estimates?) in question: Impact of recession on land rent responses Impact of planting boom in NRI dataset on empirical rent response Silvicultural response of non-integrated forest landowners Wedge between timber rents and timberland prices Dominance of PST as rent/landuse driver Inelastic Supply-Price elasticity as owners wait for price recovery 09/27/2016 GHG Sherphardstown 37

38 QUESTIONS? 09/27/2016 GHG Sherphardstown 38

39 No Analog States Modeling implications If its only data, then we can likely use the same models If the data generating processes (production functions or market relationships) have changed, then the models based on these processes may not perform well. Could be as simple as changing parameters (eg the elasticity of price wrt timber inventory), but could also challenge the fundamental basis for our models (eg profit maximization under uncertainty) 09/27/2016 GHG Sherphardstown 39

40 To go where no one has gone before Introduction: Are we in a place where no one has gone before in southern timber markets? No analog states defn Includes situations where the data are outside the HRV And where the processes have changed Modeling implications If its only data, then we can likely use the same models If the data generating processes (production functions) have changed, then the models will need to be changed. Could be as simple as changing parameters (eg the elasticity of price wrt timber inventory), but could also challenge the fundamental basis for our models (eg profit maximization) Define timber products, bmts, growth/removals/inventory, growth/removals ratio?? Implications of NAS in timber markets matter alot history matters and market linkages matter Cant increase harvestable pulpwood inventory over the next 10 years or redefine pulpwood Only way to increase sawtimber inventory over the next 20 years is to decrease the harvest of pulpwood over the next 20 years or redefine sawtimber Timber products are linked through time, through harvesting (come along volume or cull), and through residues Are we currently in NAS? What are we observing that makes us think yes? And what are the implications of each? What is happening now southwide? Demand. Supply. Prices. Pw=.3*sawt Land use. Stable but not static Are we currently in NAS? What are we observing that makes us think yes? And what are the implications of each? (SE GA examples) Changes in demand. Recession led to decline in sawt demand. Increases in wood energy and osb demand. Changes in supply. All that planting in the 90s came of age and we have sawtimber overhang. New pw demands, at same time as decreased harvest, led to increases in pw harvest. Different Age Class Distribution than in past. These changes lead to changes in prices. which leads to reduced rents and thus a change in land in forestry. What kinds of changes in processes and what are implications and can we model them? Landowner responsiveness to changes in rents. Could become less predictable. Could increase ie landowners get tired of waiting for price comeback and start harvesting at any little increase in price. This would keep prices down, would reduce rents, leads to more harvest, but also less timberland. Complicated. Yes we can model. There must be other examples. Conclusions. Changes in demand. Recession led to decline in sawt demand. Increases in wood energy and osb demand. 09/27/2016 GHG Sherphardstown 40

41 So Will Landowners Move to Short NOT Likely: 1.5 Price Ratio + Doubling of Volume in 10 Years = 3.5 x Value Increase = 13% IRR BUT: In the Short Run: Small Sawtimber (CNS) Goes to Pulp Mills Rotation Plantation Mgt? 09/27/2016 GHG Sherphardstown 41

42 What Happens to Prices when a Long Wave Inventory Cycle Hits a Down Housing Cycle? ALFLGA w/inventory effect (WoW affects price and removals) ALFLGA no inventory effect (what if the WoW didn t affect price) PST prices slow rebound PST prices rebound 09/27/2016 GHG Sherphardstown 42

43 UP 10%-30% LESS THAN 10% CHG Down 10%-30% What does the latest TPO Data Show? Pine PULP consumption UP (fluff and mill residue) Pine SAW consumption DOWN 09/27/2016 GHG Sherphardstown 43

44 Inventory by dbh class South Carolina (index) Georgia (index) Product equilibrium requires small roundwood to expand from 9 to 13 dbh 09/27/2016 GHG Sherphardstown 44

45 How does Pine Sawtimber Inventory Affect Stumpage Prices? ALFLGA pine sawtimber South pine sawtimber 09/27/2016 GHG Sherphardstown 45

46 Prices through te Recession and Recovery timber-mart south PST Stumpage Prices Have Not Recovered 09/27/2016 GHG Sherphardstown 46

47 SE-GA Growth vs Removal Trends long wave cycles 09/27/2016 GHG Sherphardstown 47