Towards More Accurate Growth Simulations and NPV Appraisals:
Using INFORM to Project Tree Grade and Market Value Increases

Published in The Compiler, Spring 1996 (Vol. 14, No. 1)

Common Appraisal Methods

Appraisals of forest land are usually required for purposes of sale or purchase, for estate valuations and for long term financial planning.  Real estate appraisers often use the technique of market data valuation (American Institute of Real Estate Appraisers 1978), whereby information is gathered on recent sales of tracts of land with acreages and development potentials similar to the tract in question.  The development potentials of tracts to be compared are assigned values based on the number of potential house lots at whatever the market rate is in the area.  These houselot values are then subtracted from the total sale values to find the forest land values.

The market data valuation method will usually indicate widely varying values for forest land, and is influenced by many factors other than the present and future potential for forestry use.  These other factors may include scenic qualities, proximity to centers of population, and other qualities unique to the particular tracts under consideration.  This appraisal method ignores the very different values of  timber and growing stock on different tracts of land.

A variation of the market data approach is currently being used by the Massachusetts Division of Fisheries and Wildlife in its land acquisition activities (Scanlon, personal communication 1995).  First, land and growing stock are valued at $200 to $300 per acre.  The range allows for some variation in site quality and accessibility and is based mainly on current market data valuations.  Second, marketable timber is valued at the current market rates for the particular species, sizes, and qualities on the property.  The total value is obtained by adding the land and growing stock value to the marketable timber value.

Financial Terminology Sidebar A

Net Present Value Method

More sophisticated techniques of forest appraisal have been developed by forest economists (Duerr 1960, Bullard and Straka 1993).  These net present value techniques involve projections of future timber volume and value increases and calculations of present value based on the cost of investing ones capital in the trees and land for given periods of time.  Net present value appraisals portray the value of forest investments more accurately than do current value appraisals.  This is especially true with timber and growing stock that will grow at higher rates than the alternative or discount rate which is used to calculate the cost of investing ones capital (Nodine 1996 in review). 

Projections of future timber volume and value can be made with the help of computer growth simulation programs.  INFORM is one of several programs capable of simulating all the ways in which trees grow in volume and value (Hepp 1992).  INFORM is also capable of calculating net present value and other measures of financial profitability.

Like most simulators, INFORM has sophisticated volume growth algorithms based on analysis of extensive forest inventory data from different parts of the country.  They will predict volume growth and mortality for different species on different sites and with different levels of competition from other trees.  These algorithms may be set as defaults for different geographic regions.  However, INFORM does not include default algorithms for grade and market value increases for different species on different sites. 

INFORM permits grading of the individual trees that comprise the inventory data.  INFORMs growth simulation program, YIELD-MS, will advance trees to higher grades as they meet the diameter requirements for those grades (Hepp 1992). YIELD-MS includes one set of grade increase probabilities for all species and sites.  These probabilities are based on limited data.  Most users will want to enter their own probabilities for different species and sites.

INFORM also allows users to enter rates of market value increase for different species and grades.  This is an important feature where market values tend to increase at very different rates for different species and grades (Davies 1991, Nolley 1994, Luppold and Baumgrass 1995).  Most users will want to enter their own estimates for future market value increases.  Together, future grade and market value projections are very important for accurate net present value appraisals.

Tree Grade Specifications

The Forest Service hardwood tree grading system (Hanks 1976) is widely recognized but little utilized.  Despite efforts to make it easier to apply (Miller et al.1986, Liu and McLaren 1989), most foresters have developed their own grading systems because they find that the Forest Service system is both too complicated and too irrelevent to tree value, particularly for larger trees.  The Forest Service tree grading system does not work easily with INFORM because it uses three inch diameter classes while INFORM requires even number diameter classes.  The Forest Service tree grading system cannot be directly related to commonly used sawmill log grading systems. 

Softwood Tree Grades Sidebar B

These problems indicate the need for a simple, practical tree grading system that can be used with INFORM and that relates to sawmill log grading systems.  This is especially the case with high grade hardwoods.  A consensus hardwood log grading system is evolving in southern New England.  There is variation among sawmills in their specifications, but the differences are minor.  Table 1 shows average hardwood log grade specifications and prices for red oak logs in southern New England in spring of 1995.  The specifications and prices are from an informal sampling of sawmills within a 50 mile radius of Northampton, MA (after Doolan 1995).

Log Grade	MinimumTip Diameter	Minimum Length	Clear Faces	Mill $ per Mbf	Stump $ per Mbf
Prime	16"	8' or 10'	4	880	760
Select	14"	8'	4	720	600
One	12"	8'	3	560	440
Two	12"	8'	2	340	220
Three	10"	8'	1	180	60
Four	10"	8'	0	100	-20

Table 1. Hardwood log grades with average specifications and prices for red oak in southern New England, spring 1995.  Stumpage prices assume $120 per Mbf for logging and trucking.

Log grades are determined by even number tip diameters and number of clear faces.  Specifications are generally the same for all species.  A simple, practical tree grading system should be based on the same specifications.  Table 2 shows such a tree grading system for hardwoods.

Tree Grade	Minimum DBH	Minimum Height, Tip D	Quality Requirements
1	18"	16'/10"	4 clear faces
2	16"	16'/10"	3 clear faces
3	14"	16'/10"	2 clear faces
4	12"	8'/10"	1 clear face
5	10"	8'/10"	sound

Table 2.  A provisional hardwood tree grading system derived from log grades described in Table 1.

In this system tree grade specifications correspond to log grade specifications of the second 8 log.  For example, a grade 1, form class 78, 18 DBH tree will have a 16 tip 8 prime butt log and a 14 tip 8 select second log.  Tip diameter and clear face specifications correspond as follows: tree grade 1 ~ log grade select, tree grade 2 ~ log grade 1, tree grade 3 ~ log grade 2, tree grade 4 ~ log grade 3, tree grade 5 ~ log grade 4.  Per Mbf tree values may be estimated as a percentage of the per Mbf price of the second 8 log.  Table 3 shows a simple method for estimating tree value according to log grades, volumes and prices.

8' Log	Tip Diameter	Log Grade	bf Intnl*	$ per Mbf	Log $
1	16"	Prime	85	760	65
2	14"	Select	65	600	39
3	12"	One	45	440	20
4	10"	Two	30	220	7
Totals:			225		130
$ per Mbf  for Whole Tree:				578
Whole Tree $ per Mbf  / Second 8 Log  $ per Mbf:				96%
*Ashley 1989

Table 3.  Estimated tree value for a red oak, tree grade 1, 18 DBH, 2 logs using log grade specifications and prices from Table 1.

Unless trees have serious defects, they will grow into the next higher grade in about 10 years at 10 rings per inch.  As the stumpage prices in Table 1 indicate,  red oak values increase by roughly 25 to 300% with each increase in log grade.  The same is true for most other high grade hardwood species. These grade value increases translate into real (after subtracting the rate of inflation) compound rates of increase of 3 to 12% per year for high grade species (Lundgren 1971).

Probabilities for Tree Grade Increases

Not all trees are capable of growing into the next higher grade because of  branches, knots, seams or sweep (Ernst and Marquis 1979, Yaussey 1993).  This is particularly true of poorer quality trees which are typically marked in commercial thinnings, improvement harvests and shelterwood cuts.  Under these silvicultural treatments, residual trees are more likely to increase in tree grade.  Table 3 shows ranges of grade increase probability for residual hardwood trees derived from field inventories in western Massachusetts in which trees were graded for present and estimated 15 year potential grade.  The ranges allow for differences in species and site quality.

Grade Path:	5 to 4	4 to 3	3 to 2	2 to 1
Probability:	80-95%	75-90%	70-85%	65-80%

Table 4.  Ranges of tree grade increase probability for residual hardwood trees derived from field inventories in western Massachusetts.

The YIELD-MS  program in INFORM permits users to insert grade increase probabilities in the MASPEC module, along with threshold diameters for each grade.  To accomodate differences in site quality, different MASPEC. DAT files may be created with different probabilities of grade increases and renamed as necessary.  For example, files named MASPEC1, 2 or 3.DAT for high, medium and low site classes could be renamed MASPEC.DAT prior to processing inventory data.

With average hardwood growth rates of 10 rings per inch and two inch tree grades, growth simulations must be for at least 10 years to show increases in tree grade.  Three inch grading systems would require longer periods of time to show grade increases.  Growth simulators like YIELD-MS see odd number tree diameters as the next lowest even number diameter.  This means that grade increases will take even longer for odd number diameter trees.  For example, a 13 DBH Forest Service grade 2 tree would be seen as a 12 DBH tree.  It would have to grow 4 in DBH to become a Forest Service grade 1 tree.  This would require a 20 year growth simulation at 10 rings per inch.

INFORM Upgrade Sidebar C

Market Value Increase Rates

Different species shave shown very different rates of market value increase in recent years, including some very high rates for the more valuable hardwoods (Davies 1991, Nolley 1994, Luppold and Baumgrass 1995).  Realistic simulations of future value growth require estimates of rates of market value increase.  The TIMSALE module of YIELD-MS permits users to enter rates of `inflation` for each species and grade.  Inflation in this sense may be defined as the total rate of market value increase, including producer price inflation and real market value increase, or it may be defined as the real rate of market value increase alone.  In either case, users should be consistent as to which definition they use throughout YIELD-MS.    

Table 5 shows stumpage prices for various species and tree grades.  Prices were adapted from data in the Southern New England Stumpage Price Survey Results (Cooperative Extension Services of Massachusetts and Connecticut 1985-94) for the area west of the Connecticut River 1985-94.  These reports show high, median and low prices reported by foresters, loggers and sawmill owners for each quarter of the year; prices are reported separately for the areas east and west of the Connecticut River.  High, median and low prices in the stumpage reports were adapted to the grading system in Table 2 as follows: the high price is equivalent to just below grade 1, the median price is equivalent to just below grade 2, the low price is between grades 3 and 4.

      

Species, Grade	1985	1986	1987	1988	1989	1990	1991	1992	1993	1994
Red Oak
1	400	450	700	550	560	600	600	610	620	680
2	180	220	400	300	320	300	300	370	500	450
3	100	110	200	150	200	150	200	250	240	200
4	50	60	100	80	100	70	100	100	160	100
Ash, Cherry
1	200	300	420	500	400	280	300	400	450	540
2	120	160	210	280	220	150	140	160	200	280
3	70	80	100	120	100	90	80	80	120	200
4	40	40	50	40	50	40	40	40	60	100
Hard Maple
1	100	150	120	130	160	140	220	300	500	460
2	70	90	80	70	80	80	100	160	240	230
3	50	60	50	40	40	40	50	70	100	80
4	30	30	30	20	20	20	30	40	40	40
Black/Yellow Birch
1	80	80	100	110	120	120	140	160	180	200
2	50	50	60	70	70	70	80	100	120	100
3	30	30	40	40	40	40	50	60	70	70
4	20	20	20	20	20	20	30	30	30	30
Beech, Red Maple
1	70	60	70	90	80	100	90	80	80	100
2	50	50	50	60	60	60	50	60	60	60
3	30	40	40	30	40	30	30	40	40	40
4	20	30	30	20	20	20	20	20	20	20
White Pine
1	100	90	110	100	100	110	80	90	100	130
2	70	70	70	60	70	70	60	60	70	70
3	50	50	50	40	40	40	40	40	50	50
4	30	30	30	20	20	20	20	20	30	30
Red Pine
1	60	60	70	60	70	70	70	80	100	110
2	40	40	50	50	60	50	50	50	60	70
3	30	30	30	30	40	30	30	40	50	50
4	20	20	20	20	30	20	20	20	30	40
Hemlock
1	60	60	60	70	60	40	40	60	80	70
2	40	40	50	40	30	30	30	40	50	40
3	30	30	30	30	20	20	20	30	30	30
4	20	20	20	20	10	10	10	10	20	20
Wood
1	15	20	20	10	10	10	10	10	10	10
2	5	5	5	5	0	5	5	5	5	0
Pulp
1	0	0	0	5	5	5	15	10	15	5
2	0	0	0	5	5	0	5	5	5	0

Table 5.  Fall/winter stumpage prices for different species and tree grades as adapted from the Southern New England Stumpage Price Survey Results 1985-94 for the area west of the Connecticut River Prices adapted to fit tree grading system in Table 2.  Timber prices rounded off to the nearest $10 per Mbf.

In order to calculate the rate of market value increase for different species and grades, the data for each row in Table 5 were first subjected to regression analysis in the spreadsheet program Lotus 1-2-3.  The smoothed results for even number years are shown in Table 6.  The ranges of values derived from this process were then subjected by internal rate of return analysis in Lotus 1-2-3 to determine the annual rate of market value increase.  A 3% rate of producer price inflation (U.S. Department of Commerce 1995) was subtracted from these nominal rates to determine real rates.                               

Species, Grade	1986	1988	1990	1992	1994	Rate	Real Rate
Red Oak
1	480	520	560	600	640	3%	0%
2	220	270	320	370	420	7%	4%
3	120	150	180	210	240	8%	5%
4	60	75	90	105	120	8%	5%
Ash, Cherry
1	290	330	370	410	450	5%	2%
2	165	180	195	210	225	4%	1%
3	70	85	90	105	120	6%	3%
4	30	40	50	60	70	10%	7%
Hard Maple
1	180	250	320	390	460	11%	8%
2	70	110	150	190	230	14%	11%
3	40	50	60	70	80	8%	5%
4	20	25	30	35	40	8%	5%
Black/Yellow Birch
1	70	100	130	160	190	12%	9%
2	45	60	75	90	105	10%	7%
3	25	35	45	55	65	11%	8%
4	15	20	25	30	35	10%	7%
Beech, Red Maple
1	70	75	80	85	90	3%	0%
2	50	55	55	60	60	2%	-1%
3	35	35	40	40	40	1%	-2%
4	20	20	20	20	20	0%	-3%
White Pine
1	100	105	110	115	120	2%	-1%
2	70	75	75	80	80	1%	-2%
3	50	50	50	50	50	0%	-3%
4	30	30	30	30	30	0%	-3%
Red Pine
1	60	70	80	90	100	6%	3%
2	40	45	50	55	60	5%	2%
3	30	30	35	40	45	5%	2%
4	20	20	25	30	30	5%	2%
Hemlock
1	55	60	65	70	75	4%	1%
2	40	40	40	45	45	1%	-2%
3	25	25	25	30	30	2%	-1%
4	15	15	15	20	20	3%	0%
Wood
1	10	10	10	10	10	0%	-3%
2	5	5	5	5	5	0%	-3%
Pulp
1	0	5	5	10	10	N/A	N/A
2	0	0	0	5	5	N/A	N/A

Table 6.  Stumpage prices for different species and tree grades as adapted from the Southern New England Stumpage Price Survey Results 1985-94 for the area west of the Connecticut River after regression analysis and data smoothing.  Rate = nominal annual interest rate necessary for prices to increase as indicated 1986-94.  Real rate = nominal rate less 3% annual producer price inflation (U.S. Department of Commerce 1995). Timber prices are rounded off to the nearest $5 per Mbf.

This information is useful in comparing the investment values of different species and grades over the past 10 years, and in estimating rates of market value increase for growth simulations.  Any projections of future market value increases should be carefully qualified as to the assumptions that underly them.  This information can also be used in simulating the value growth of old inventories to the present with INFORM's BATCH YIELD-MS program.  In this way old inventory data can be accurately updated to current volumes and values for appraisal purposes. 

Grade Value Increases for Large Trees Sidebar D

Stumpage price reports have been available for most states for enough years (Lutz et al. 1992, Rosen and Carryl 1993, Sendak 1994) to make this type of analysis possible almost anywhere.  However, many reports do not include high, median and low values from which tree grade values may be extrapolated.  Values for different tree grades may be determined by gathering log price data from local sawmills (Doolan 1996) and subtracting logging and trucking costs as shown in Tables 1 and 3.  There may be some question as to how closely stumpage values for log grades correspond to stumpage values for tree grades.  Reconciling these differences may best be left to trial and error.

Discussion

Table 7 illustrates the differences in internal rate of return (IRR) and net present value (NPV) for YIELD-MS growth and treatment simulations for the same stand with and without grade and market value increases.  Actual inventory data for a managed stand in Chesterfield, MA were grown for 15 years with YIELD-MS and then (theoretically) clearcut to make YIELD-MS perform profitability analyses.

Measure	Volume Increase Only	Volume and Market Value Increase	Volume, Grade and Market Value Increases
Initial $ per Acre	481	481	481
15 Year $ per Acre	1216	1216*	2185*
IRR %	5.0	10.3	14.6
NPV $	-60	371	1019

*The current version of YIELD-MS (3.0) does not show market value increase (`inflation`)  in the values of growth simulations, although it does account for market value increase it its IRR and NPV calculations.  Therefore these values are low, but the IRR and NPV values are correct.

Table 7.  Per acre financial analyses for YIELD-MS simulations of growth and treatment of a 60-year-old mixed hardwood stand after a commercial thinning to release the better trees.  Site index for red oak is 65; predominant residual species are red oak, yellow birch and white ash.  Tree grade specifications and prices were taken from Tables 2 and 6; market value increase rates were adapted from Table 6 real rates.  Investment basis was the current clearcut value of the timber and growing stock; real discount rate was 6%.  Costs included foresters fees and taxes.

Net present value appraisals will generally show higher values than appraisals based on market data.  This is particularly the case with managed stands of high grade hardwoods.  The difference between the results of the different appraisal methods may be attributed to 1) lack of public awareness of the value of timber investments and 2) more severe discounting of forest investments due to their lack of liquidity and the inconveniences of land ownership (Rose et al. 1988).  Table 8 compares the results of the different appraisal techniques for the stand used in creating Table 7.

Appraisal Method	Land Only $ per Acre	Current Timber $ per Acre	NPV @ 6% $ per Acre	Total $ per Acre
Current Market Data	500	N/A	N/A	500
Current Land and Timber	250	440	N/A	690
Full NPV Timber and Growing Stock	41 (Growing Stock)	440	1019	1500

Further Documentation Sidebar E

Table 8.  Per acre appraisals of the same forest stand using three different appraisal methods. Stand is the same as used in Table 7 and has no real estate development potential.

Conclusions

Growth simulations and net present value appraisals are more accurate when tree grade and market value projections are included in the programming.  INFORMs growth simulator, YIELD-MS, is capable of projecting tree grade and market value increases into the future.  YIELD-MS can also perform financial analyses of growth simulations.  When simulations include volume, grade value and market value increases, total real value increases may be 2-3 times volume increases alone.  Net present values will be correspondingly higher.  Net present value appraisals will generally show higher values than market data appraisals, particularly for high quality hardwood stands that have little or no potential for real estate development.

Disclaimer

The log specification and price information in Table 1 is from a limited sample.  For this reason, the grading system in Table 2 and the valuation method in Table 3 should be viewed as only provisional.  Likewise, the market value trends and rates of market value increase in Table 6 are derived from a limited data source. Projecting these rates into the future assumes that market dynamics will be the same as they have been in southern New England over the past 10 years.  This assumption may be unwarranted, particularly for other regions.

References

American Institute of Real Estate Appraisers. 1978. The Appraisal of Real Estate. Chicago.

Ashley, B.S. 1989. Reference Handbook for Foresters.  USDA For. Serv. NA-FR-2. U.S. Government Printing Office. Washington, DC.

Bullard, S.H. and T.J. Straka. 1993. Basic Concepts in Forest Valuation and Investment Analysis.  GTR Printing.  Starkville, MS.

Cooperative Extension Services of Massachusetts and Connecticut. 1985-94. Southern New England Stumpage Price Survey Results. Mass. Dept. Env. Mgt.  Boston.

Davies, K. 1991. Forest investment considerations for planning thinnings and harvests. North J. Appl. For. 8(3):129-131.

Doolan, R. 1990-96. Sawlog Bulletin. Littleton, NH.

Duerr, W.A. 1960.  Fundamentals of Forestry Economics. McGraw-Hill. New York.

Ernst, R.L. and D.A. Marquis. 1979. Tree grade distribution in Allegheny hardwoods. USDA For. Serv. Res. Note NE-275.

Hanks, L.F. 1976. Hardwood tree grades for factory lumber. USDA For. Serv. Res. Pap. NE-333.

Hepp, T. 1992.  INFORM Users Manual. Tennessee Valley Authority. Norris, TN.

Liu, C.J. and D.J. McLaren. 1989. Use of a classification algorithm simplifies process: A model for determination of hardwood tree grade. J. of Forestry 87(4):47-48.

Lundgren, A.L. 1971. Tables of compound /discount interest rate multipliers for evaluating forestry investments. USDA For. Serv. Res. Pap. NC-51.

Luppold, W.G. and J.E. Baumgrass. 1995. Price trends and relationships for red oak and yellow poplar stumpage, sawlogs, and lumber in Ohio: 1975-1993. North. J. of Appl. For. 12(4):168-173.

Lutz, J., T.E. Howard and P.E. Sendak. 1992. Stumpage price reporting in the northern United States.  North. J. of Appl. For. 9(2):69-73.

Miller, G.W., L.F. Hanks and H.V. Wiant. 1986. A key for the Forest Service hardwood tree grades. North. J. of Appl. For. 3(1):19-22.

Nodine, S.K. 1996 in review. Discounting cash flows for forest appraisal and valuation. The Compiler 14(1).

Nolley, J.W. 1994. Bulletin of hardwood market statistics: First, second and third quarters--1994. USDA  For. Serv. Res. Note NE-359.

Rose, D.W., C.R. Blum and G.J. Brand. 1988. A guide to forestry investment analysis. USDA For. Serv. Res. Pap. NC-284.

Rosen, B.N. and N. Carryl. 1993. Price reporting for nonindustrial private owners. J. of  Forestry 91(1):34-41.

Scanlon, J. 1995. Personal communication. Mass. Div. of Fisheries and Wildlife. Westborough, MA.

Sendak, P.E. 1994. Northeastern regional timber stumpage prices: 1961-1991. USDA For. Serv. Res. Pap. NE-683.

U.S. Department of Commerce. 1995. Statistical Abstract of the United States . U.S. Government Printing Office. Washington, DC.

Yaussey, D.A. 1993. Method of estimating potential tree-grade distributions for northeastern forest species. USDA For. Serv. Res. Pap. NE-670.

Appendix A: Provisional Softwood Tree Grading Systems

White Pine

Grade*	Minimum DBH	Minimum HT/Tip D	Quality Requirements
1	18"	24'/10"	knots <1" D
2	16"	24'/10"	knots <2" D
3	14"	16"/10"	knots <3" D
4	12"	16'/10"	knots <4" D
5	10"	8'/10"	knots <5" D
*Specifications for red-knotted trees. Drop one grade for black-knotted trees.
Hemlock
Grade	Minimum DBH	Minimum HT/Tip D	Quality Requirements
1	18"	48'/10"	sound, FC 75
2	16"	32'/10"	sound, FC 75
3	14"	16'/10"	sound, FC 75
4 5	12" 10"	16'/10" 8'/10"	sound, FC 65 sound, FC 65
Red Pine
Grade	Minimum DBH	Minimum HT/Tip D	Quality Requirements
1	14"	48'/8"	straight, knots < 2" D
2	12"	32'/8"	straight, knots < 2" D
3	10"	16'/8"	straight, knots < 2" D
4 5	10" 10"	16'/8" 8'/8"	sound sound