COURT OF APPEALS OF VIRGINIA
Present: Judges Coleman, Willis and Bumgardner
Argued at Richmond, Virginia
GEORGE C. HOWELL, III
OPINION BY
v. Record No. 2800-98-2 JUDGE RUDOLPH BUMGARDNER, III
JANUARY 27, 2000
MARGARET H. HOWELL
FROM THE CIRCUIT COURT OF THE CITY OF RICHMOND
Theodore J. Markow, Judge
Ronald S. Evans (Lawrence D. Diehl; Brenner,
Dohnal, Evans & Yoffy, P.C., on briefs), for
appellant.
James C. Roberts (Dawn B. DeBoer; Mays &
Valentine, L.L.P., on brief), for appellee.
George C. Howell, III appeals from the final decree
divorcing Margaret H. Howell and him and distributing their
marital estate. He complains primarily that the trial court
erred in valuing his interest in his law firm, but he also
asserts it erred in classifying a money market account, in
allocating marital debt, and in setting spousal support. The
trial court referred all issues to a commissioner in chancery
and adopted nearly all of the commissioner's findings in its
final decree. Concluding that the trial court did not err, we
affirm.
The parties married in 1982 after the husband graduated
from law school but before he began practicing law with Hunton &
Williams, a large firm based in Richmond. The husband
specialized in tax law, established a successful practice with
the firm, became a partner in 1989, and earned an income of more
than $400,000 by 1998. After the first of their two children
was born in 1984, the wife no longer worked outside the home.
She took care of their two children and maintained the
household. When the parties separated November 17, 1995, she
maintained custody of the children, and when they divorced
November 4, 1998, she received sole physical custody of the
children with the husband's consent.
The husband conceded that his partnership interest was
marital property because he acquired it during the marriage.
The partnership agreement at Hunton & Williams defined the value
of the partnership interest upon termination or death. It
entitled the partner to receive only the balance of his capital
account and his share of the net income. As applied to this
case, those two items amounted to $85,614. The husband
maintains that the agreement fixed the value of his partnership
interest for equitable distribution purposes and that it
precluded consideration of whether his interest had goodwill.
He argues that Kaufman v. Kaufman, 7 Va. App. 488, 375 S.E.2d
374 (1988), ruled that restrictive agreements control the
valuation of business interests and limit the value to that
established by the agreement. The husband argues that the trial
court erred in finding that his partnership interest had
- 2 -
goodwill because the Hunton & Williams agreement made no
provision for goodwill.
Code § 20-107.3(A) directs that the trial court value all
property of the parties, but it does not define the term,
"value," for equitable distribution purposes. The statute does
not set the standard of value, that is, the measure of
property's worth for equitable distribution. "Value" is a
mercurial term; the term has numerous, distinct meanings. The
various meanings are not interchangeable. The meaning of the
term, "value," depends on what is being valued, who is
interested, and why it is being valued. A piece of property may
have different values for different purposes. The purpose for
which it is being valued determines which definition, which
standard of value, is proper. Purpose determines the standard
of value; that, in turn, determines the appropriate methods of
valuation.
Bosserman v. Bosserman, 9 Va. App. 1, 384 S.E.2d 104
(1989), defined "value" for equitable distribution purposes; it
set intrinsic value as the standard of value. "Trial courts
valuing marital property for the purpose of making a monetary
award must determine from the evidence that value which
represents the property's intrinsic worth to the parties
. . . ." Id. at 6, 384 S.E.2d at 107. The value of an item of
marital property is its intrinsic worth to the parties: the
- 3 -
worth to the husband and wife, the parties; the value to the
marital partnership that the court is dissolving.
Intrinsic value is a very subjective concept that looks to
the worth of the property to the parties. The methods of
valuation must take into consideration the parties themselves
and the different situations in which they exist. The item may
have no established market value, and neither party may
contemplate selling the item; indeed, sale may be restricted or
forbidden. Commonly, one party will continue to enjoy the
benefits of the property while the other must relinquish all
future benefits. Still, its intrinsic value must be translated
into a monetary amount. The parties must rely on accepted
methods of valuation, but the particular method of valuing and
the precise application of that method to the singular facts of
the case must vary with the myriad situations that exist among
married couples.
Because intrinsic value must depend on the facts of the
case, we give great weight to the findings of the trial court.
We affirm if the evidence supports the findings and if the trial
court finds a reasonable evaluation based on proven methodology
and on the application of it to the particular facts of the
case. See Russell v. Russell, 11 Va. App. 411, 415-16, 399
S.E.2d 166, 168 (1990). "[T]he trial court's valuation of
goodwill will not be disturbed if it appears that the court made
a reasonable approximation of the goodwill value, if any, of the
- 4 -
professional practice based on competent evidence and the use of
a sound method supported by that evidence." Id. at 417, 399
S.E.2d at 169 (citation omitted).
In Kaufman, the husband purchased a one-third interest in a
medical practice six months after the parties separated. The
purchase contract required stockholders to resell their interest
to the corporation at a value excluding any provision for
accounts receivable or work in progress. The trial court held
that the value set by the contract controlled because the terms
were made at arm's length and were devoid of fraud. This Court
affirmed that valuation and ruled it was bound by the finding
because there was evidence to support it. See 7 Va. App. at
502, 375 S.E.2d at 381 (citing Code § 8.01-680).
In Bosserman, the wife asked the court to value the
husband's interest in a family-owned, closely-held corporation.
A restrictive agreement required a shareholder to offer the
stock first to the corporation at "true book value." The
husband argued that the restriction defined the value. See 9
Va. App. at 3, 384 S.E.2d at 106. This Court recognized that
the value set by such agreements "is often artificial and does
not always reflect true value," even though the agreements may
be binding on business partners. Id. at 6, 384 S.E.2d at 108.
"[T]he sale price set by restrictive provisions on transfer is
not conclusive as to the value of the stock." Id. at 7, 384
S.E.2d at 108. A restriction on transfer does not control its
- 5 -
value, "but the restriction on transfer is a factor which
affects the value of the stock for purposes of equitable
distribution." Id. (citations omitted). Interestingly,
Bosserman relied upon an article discussing goodwill valuation
in law firms to conclude restrictive provisions are not
conclusive of value.
Russell established that a psychiatric practice could have
goodwill that had value. "[I]f the trial court determines from
the greater weight of the evidence that a professional practice
has goodwill value, that amount must be subject to valuation as
part of the marital property. To hold otherwise would result in
a windfall to the professional spouse." 11 Va. App. at 416, 399
S.E.2d 168. Goodwill may be an asset of a professional
practice, and if it is, it is subject to valuation for equitable
distribution purposes.
All three decisions turned upon the particular findings
made from the precise evidence presented to the trial court.
The holdings establish that the value of property is an issue of
fact, not of law. Goodwill may exist, or may not; an agreement
may restrict its value, or may not. The existence of goodwill
must be proven, and if it exists, its value must be proven. The
evidence presented at trial determines the result, and the
result may vary from case to case as the evidence differs. If
conflicting, competent evidence is presented, that found more
credible will determine whether goodwill exists and its value if
- 6 -
it does exist. The existence is not fixed as a matter of law,
nor is the method of valuation; both are functions of the fact
finding process.
In this case, the commissioner found that the husband's
partnership interest had intangible value, goodwill, and that
the partnership agreement did not reflect its true worth to the
parties, its intrinsic value. After hearing extensive expert
testimony, the commissioner accepted the opinion of the wife's
expert and valued the partnership interest at $319,569. The
trial court accepted the commissioner's findings.
The wife's certified public accountant testified that the
husband's interest in his law firm had value in the nature of
goodwill. The husband's expert witness, also a certified public
accountant, disagreed and testified that the partnership
agreement controlled. Both opinions were made by qualified and
experienced experts, were reasoned and based on accepted
principles, and either might have been found credible by the
trial court. In resolving the conflicting opinions, the
commissioner gave greater weight to the wife's expert, and the
trial court accepted that finding.
The trial court has discretion to resolve conflicting
expert testimony to determine an asset's value. See Rowe v.
Rowe, 24 Va. App. 123, 140, 480 S.E.2d 760, 768 (1997); McDavid
v. McDavid, 19 Va. App. 406, 413, 451 S.E.2d 713, 718 (1994).
Where the court accepts the commissioner's findings of fact,
- 7 -
"the findings are presumed to be correct when reviewed on appeal
and are to be given 'great weight' by this Court." Rowe, 24 Va.
App. at 140, 480 S.E.2d at 768 (citation omitted). The court's
decision will not be disturbed on appeal unless plainly wrong or
without evidence to support it. See Blank v. Blank, 10 Va. App.
1, 9, 389 S.E.2d 723, 727 (1990).
The standard of value has a critical impact on the effect
of a restrictive agreement. The standard in equitable
distribution, intrinsic value, looks to the value to the parties
to the divorce action. When the parties do not contemplate sale
and one of the two is going to enjoy the benefits of the
property with no likelihood of leaving the business, a
restrictive agreement may have less bearing on the value to the
parties. The agreement should be considered but it is not
conclusive. "The reason for rejecting the value set by buyout
provisions is that they do not necessarily represent the
intrinsic worth of the stock to the parties." Bosserman, 9 Va.
App. at 6, 384 S.E.2d at 107.
The commissioner considered the restrictive agreement as
one factor in determining whether goodwill existed, but he did
not find it to be the sole, controlling factor. Credible
evidence supported the finding that the partnership agreement
was but one factor to consider in determining the intrinsic
value of the husband's partnership interest. The trial court
did not err in finding that the Hunton & Williams partnership
- 8 -
agreement did not control the existence of goodwill in the
husband's interest in the firm.
The husband contends that even if the trial court did not
err in rejecting the partnership agreement valuation, it did err
in the method and application used to value the interest. Both
sides presented expert testimony valuing the goodwill of the
partnership interest. The commissioner found that the wife's
expert was persuasive and accepted his evaluation. The trial
court affirmed those findings fixing the value at $319,569.
The wife's expert used the excess earnings method to value
the practice goodwill, a method acknowledged by the husband's
expert as the most popular method for evaluating professional
goodwill. The wife's expert began by comparing the husband's
average income for three years to that of a peer group. The
difference between the two was the excess earnings due to the
husband's association with the firm, but not the earnings due to
his personal efforts. It was the additional income he received
from being associated with his law firm. The wife's expert
projected those excess earnings over the husband's expected
career with Hunton & Williams. Using the discounted future
earnings method, he then calculated the present value of the
husband's total future excess earnings.
The wife's expert began by comparing the husband's earnings
to those of a peer group. The husband argues that this method,
described as working from the "bottom up," is incorrect. He
- 9 -
contends that the peer group selected by the wife's expert
compared him to merely average attorneys, did not reflect his
special reputation and expertise, and failed to distinguish his
personal credentials from those of Hunton & Williams. He also
contends that the discounted future earnings method is an
improper method of valuation because it is based on future
earnings projections. Finally, he argues he can have no excess
earnings due to his association with Hunton & Williams because
he had received offers from other firms for 150-175% more
income.
The wife's expert carefully explained why he selected the
methodology that he used. A major reason was the unavailability
of partnership financial data. Before getting the appraisal,
the wife attempted to discover the partnership's financial
records. Hunton & Williams successfully had the request quashed
and did not make any partnership financial data available. The
wife's expert felt the "bottom up" method was more accurate
under those circumstances. The trial court accepted that
opinion and analysis. We cannot say that the opinion was
unworthy of belief as a matter of law.
The wife's expert selected the husband's peer group based
on factors that included comparable geographic location,
population, firm size, specialization, and years of practice.
The husband argued that the group did not reflect his special
talents and that the appropriate peer group was the one selected
- 10 -
by his expert. The trial court carefully examined these
conflicting opinions and rejected the one offered by the
husband. The trial court found the peer group suggested by the
husband's expert was flawed because the group only included
firms located in very large cities, such as New York or San
Francisco, and it was a selection provided by the husband.
The wife's expert valued the husband's excess earnings due
solely to his association with his firm at $38,269. He then
projected this figure to the husband's expected retirement year,
2021. After adding the terminal value of the husband's capital
account, he applied a discount rate of 6.9%. The result was the
present value of the husband's partnership interest, $319,569.
The court accepted the wife's expert's use of a discount rate of
only 6.9%. The expert justified that rate, one applicable to
nearly risk-free investments, and the record does not suggest
that it was incredible. As presented, the proper discount rate
was an expert opinion that could be accepted if the trier of
fact found it persuasive and believable.
Discounting future earnings is not an inherently flawed
method of valuation because it is based on projected future
earnings. The value of goodwill can have two components.
Professional goodwill (also designated as individual, personal,
or separate goodwill) is attributable to the individual and is
categorized as separate property in a divorce action. Practice
goodwill (also designated as business or commercial goodwill) is
- 11 -
attributable to the business entity, the professional firm, and
may be marital property. The commissioner and the trial court
carefully distinguished between these two components and
selected a value that was solely attributable to the husband
being a partner in Hunton & Williams. It represented the
premium due to the husband's association with Hunton & Williams,
the economic advantage he enjoyed because he was a partner in
that firm. It included no value attributable to him personally,
and it did not rely upon any earnings due to the husband's own
expertise, reputation, experience, skill, knowledge, or
personality. As applied, the discounted future earnings method
was not a flawed method of valuation.
In valuing the goodwill of the partnership interest, courts
must take special care not to confuse the owner spouse's
personal future earning capacity with practice goodwill
attributable to the law firm in order to avoid double counting.
"Further, particular care must be given that future earnings
capacity and reputation not be confused with professional
goodwill." Russell, 11 Va. App. at 417, 399 S.E.2d at 168.
Though the husband's expert first stated that in his
opinion the partnership interest could have no goodwill value,
he offered an alternative opinion that recognized the husband's
interest did have goodwill value. The trial court found that
the alternative opinions offered by the husband, that his
interest did and did not have goodwill value, were contradictory
- 12 -
and confusing. The husband's expert used the capitalized
historic earning method. His method required comparison of
earnings to those of a peer group. The trial court found the
husband's peer group inherently flawed, and it noted the group
data was for the wrong year and compared group data to a single
year of the husband's income.
The husband's expert calculated the value of a theoretical,
marketable, controlling partnership interest in Hunton &
Williams. To that figure, the expert applied a discount of 40%
for lack of marketability and 30% for minority status. That
reduced the market value of the husband's interest to $105,177.
The trial court rejected the husband's reasoning for discounting
for minority status and marketability. It found the discounts
inappropriate because no transfer of the partnership interest
was foreseeable and no one in the firm, nor any group within it,
exercised majority control. See Bosserman, 9 Va. App. at 7-9,
384 S.E.2d at 108-09 (discussing restriction on marketability
and minority shareholder status as affecting value to owning
spouse). The trial court also rejected the argument that offers
of employment at substantially higher compensation meant all
goodwill was personal. The trial court would not accept the
conclusion, suggested by that argument, that Hunton & Williams
had no practice goodwill.
The experts disagreed on which of two accepted methods of
valuation to use and on the application of those methods. They
- 13 -
differed on the following: whether to value the individual
interest first; whether to capitalize historic earnings or to
discount future earnings; the proper peer group; the proper
income data; and the application of discounts. The opinions of
the experts conflicted, but neither was erroneous as a matter of
law. "Depending on the circumstance of each case, different
methods of valuation may reflect more accurately the actual
value of the stock to the shareholder." Id. at 7, 384 S.E.2d at
108. No single method for determining the value of professional
goodwill is preferred. See Russell, 11 Va. App. at 417, 399
S.E.2d at 169. We conclude that the trial court did not err in
selecting the method of valuation or in the application of it to
the facts presented in this case. The wife's valuation of the
husband's interest in his law firm was supported by substantial
and competent evidence.
The husband argues that the trial court erred in
classifying a portion of a $203,288.46 money market account as
marital property. The trial court designated $38,355 of the
fund as marital property and $164,933 as the husband's separate
property. The husband established the account in his name alone
after the parties separated. He made the first deposit in March
1996 when he received his annual partnership earnings
distribution.
Hunton & Williams operated on a fiscal year that ended in
March, and it paid the partners the balance of their earnings in
- 14 -
March of each year. The husband had to report his entire
partnership income on his individual tax return for the calendar
year in which the fiscal year ended. For example, he had to
report all income earned from April 1, 1995 to March 31, 1996 on
his individual income tax return for the calendar year 1996.
That would throw nine months' worth of earnings, the portion
earned in 1995, into the next individual income tax year, 1996.
The receipt of annual earnings in a lump at the end of the
partnership's fiscal year created an irregular cash flow. The
couple's income trailed their expenditure of it. To balance
expenditures with receipts of income, the parties borrowed
temporarily. They established an equity line of credit when
they first purchased the marital home. Throughout the marriage,
they had used that as a source of funds to pay expenses that
came due before the partnership distributed its earnings. When
they received the distribution at the end of the fiscal year,
they would pay the credit line down. They used the distribution
to repay the temporary financing incurred over the preceding
twelve months.
At the time the parties separated, the balance due on the
equity line was $126,801. The husband borrowed an additional
$70,000 after the separation, but the wife froze the account
when she learned of these draws. He then borrowed on two
additional accounts and used the proceeds to pay taxes and
tuition. The husband claimed that the funds borrowed after
- 15 -
separation were used for marital purposes. From July 1995 the
husband only paid interest on the equity line account, and after
separation, he did not curtail the account except to apply
$71,652.71 from the portion of the March 1996 distribution that
he considered to be marital property.
The fiscal year-end distribution that husband received in
March 1996 was for work done both before and after the date of
separation, November 1995. The husband calculated that $72,795
was marital property. He applied $71,652.71 of that to the
balance due on the equity line of credit. The remaining portion
of the March 1996 distribution became the initial deposit to his
new separate money market account. In June 1996, he received a
$50,000 bonus that also represented work done before and after
the separation. The husband calculated that $31,350 of that
bonus was earned before separation and thus was marital
property. He applied $27,000 of the marital portion toward
payment of his 1996 estimated income tax and the balance to what
he claimed were joint expenses. He deposited the portion that
he considered separate property, $18,470, to his separate money
market account.
The trial court classified $38,355 of the money market fund
as marital property because it disallowed as marital expenses
the estimated tax payment of $27,000 and $11,355 claimed as
expenses for maintaining the marital residence. The husband
argues that the use of his separate money market fund to make
- 16 -
the wife whole is inappropriate. The husband contends these
funds paid taxes, tuition, and household expenses of the marital
residence. He argues that the marital share of the June 1996
bonus paid the joint 1996 tax liability and not simply the taxes
due on the bonus.
The trial court disallowed $38,355 of his claimed marital
expenditures finding that the payment of $27,000 toward
estimated 1996 taxes was not reasonable and that $11,355 of
household expenses were not adequately proven. The court held
that the use of a substantial portion of the marital share of
the June 1996 bonus to pay taxes under the guise of a joint
obligation was not a valid marital purpose. The court also held
the husband's evidence was insufficient to satisfy his burden to
prove that the expenditures of $11,355 were for valid marital
purposes. The husband did not provide a detailed explanation of
these expenses, but the trial court still found that $54,955 was
expended for marital purposes.
The husband had the burden to establish by a preponderance
of the evidence that post-separation withdrawals of marital
funds were used for a legitimate marital purpose. See Alphin v.
Alphin, 15 Va. App. 395, 402, 424 S.E.2d 572, 576 (1992) (court
satisfied that husband's expenses, for which detailed accounting
was provided, were for valid marital purpose); Amburn v. Amburn,
13 Va. App. 661, 666-67, 414 S.E.2d 847, 850-51 (1992) (wife did
not squander away marital funds by paying for living expenses
- 17 -
husband failed to pay). We conclude that the decision to
classify a portion of the money market account to make up for
the husband's non-legitimate marital expenditures was supported
by the evidence and not plainly wrong when considering the way
the parties normally managed their finances and used the credit
line in tandem with their income.
The trial court ordered the husband to pay the balance due
on the equity line loan, on two additional bank loans, and on a
credit card account used by the wife. The equity line balance
was $122,129.50. The husband incurred the two loans of
$54,344.75 and $22,461 after the separation to pay tuition and
estimated taxes. The wife possessed only one major credit card
after the separation, but the husband stopped paying the
balance, began paying only the required minimum, and reduced the
limit to $2,500. The balance on the account was $2,191. The
husband contends that the trial court erred in allocating the
entire marital debt to him.
Throughout the marriage the parties used the equity credit
line as a source of available funds until they received the
partnership income distribution. After the separation, the
husband stopped using his income to pay recurring family
expenses. He prorated receipts, and he segregated the portion
he designated as separate property into the money market account
where it accumulated as a separate asset. He paid marital
expenses or debts incurred to pay such expenses with the portion
- 18 -
he designated as marital property. The husband used the date of
separation to divert arbitrarily his income from its established
allocation to reducing debt incurred over the previous year for
ordinary living expenses. The husband's action of not reducing
the equity line permitted the joint debt to rise while he
diverted the offsetting income to accumulate as his separate
asset. This turned marital expenses into marital liabilities of
the couple and reduced the marital wealth available for
distribution. At the same time, any income earned after
separation was diverted and not available for continuing family
obligations. The combined effect permitted the husband to
decrease the marital estate while increasing his personal
estate, to distort their financial condition to his advantage.
The trial court recognized the distorting effect upon
obtaining a correct understanding of the marital debts of the
parties. The trial court found that the equity line and credit
card account were marital debts and found the other two debts,
while incurred for marital expenses, would not have been
incurred under normal circumstances. It allocated payment of
all the debts, with the exception of the first deed of trust, to
the husband. The court carefully noted that wife could not
receive a double benefit from the equity in the home increasing
if the husband paid that debt because it excluded the equity
line debt from the calculation of the value of the marital
residence. The trial court allocated the marital debt to the
- 19 -
husband because the debt would never have been incurred if the
cycle of incurring and discharging debt, established during the
marriage, had been allowed to complete its final circuit.
Code § 20-107.3(E) empowers trial courts to distribute
marital debt. Because making an equitable distribution award is
often a difficult task, "'we rely heavily on the discretion of
the trial judge in weighing the many considerations and
circumstances that are presented in each case.'" Moran v.
Moran, 29 Va. App. 408, 417, 512 S.E.2d 834, 838 (1999)
(distribution of marital debt) (quoting Klein v. Klein, 11 Va.
App. 155, 161, 396 S.E.2d 866, 870 (1990)). Absent an abuse of
discretion, "the trial judge's determination will not be
reversed on appeal." Id.
The allocation of the debt to the husband is proper given
the manner in which this couple managed their finances. To
prevent an inequity, the trial court properly ordered the
husband to pay the debt, and it considered all the statutory
factors of Code § 20-107.3. We find no abuse of discretion.
Finally, we consider the court's award of spousal support
to the wife. The court ordered the husband to pay $7,500 per
month in spousal support and awarded child support of $2,181
based on the statutory guidelines. On appeal, he claims the
wife's expenses include some non-recurring expenses; the wife's
award should be reduced by the payment on the first deed of
trust; and that some expenses, such as mortgage payment,
- 20 -
automobile, utilities, and household expenses, should be reduced
by two-thirds because they include expenses factored into the
child support guidelines.
The wife had no independent source of income. She
submitted two exhibits to substantiate her claim for spousal
support. One listed her average monthly expenses during the
marriage ($8,746.92), and the other listed post-separation
expenses ($8,013.76). She also presented two exhibits showing
the corresponding child expenses of $3,017.23 and of $3,116.07.
The wife's post-separation expenses included non-recurring
expenses for legal bills and car rental which the court excluded
from the support award. The court declined to reduce the
support award by the amount of the mortgage payment because it
found that the wife was entitled to a reasonable housing
allowance and, when taxes on the support award were considered,
she would not have sufficient funds to cover her monthly
expenses. The court also found the husband's request to reduce
the wife's living expenses by two-thirds unreasonable in view of
her demonstrable needs. If so reduced, the wife would not be
able to meet her demonstrated financial needs, and the support
award would leave her below the standard of living to which she
was accustomed during the marriage.
The husband cites Rein v. Rein, Record No. 1120-93-1
(unpublished, Va. Ct. App. Nov. 29, 1994), as authority for
reducing these items. However, that case did not mandate such a
- 21 -
reduction. At most it directed that the trial court consider
whether there might have been double consideration of items such
as housing and utilities.
The husband is obligated to maintain the wife, within the
limits of his ability to pay, "according to the station in life
in which [she] was accustomed during the marriage." Gamble v.
Gamble, 14 Va. App. 558, 574, 421 S.E.2d 635, 644 (1992). A
spousal support award is subject to the trial court's discretion
and will not be disturbed on appeal unless plainly wrong or
without evidence to support it. See Moreno v. Moreno, 24 Va.
App. 190, 194-95, 480 S.E.2d 792, 794 (1997). Based on all the
evidence and appropriate factors, we conclude that the record
supports the spousal support award of $7,500 per month.
For the foregoing reasons, we affirm the trial court's
ruling.
Affirmed.
- 22 -