Net Commissions

Document Sample
Net Commissions Powered By Docstoc
					                  100%ers Versus Split Commission Brokers1
A farmer placed nets on his newly sown plowlands and caught a number of Cranes, which came
to pick up his seed. With them, he trapped a Stork that had fractured his leg in the net and was
earnestly beseeching the farmer to spare his life.

"Pray save me, Master," he said, "and let me go free this once. My broken limb should excite
your pity. Besides, I am no Crane, I am a stork, a bird of excellent character; and see how I love
and slave for my father and mother. Look too, at my feathers. They are not the least like those
of a crane."

The farmer laughed aloud and said, "It may be all as you say, I only know this: I have taken you
with these robbers, the cranes, and you must die in their company."

Birds of a feather flock together.



Historically, residential real estate firms have offered their brokers two differing compensation

packages. Brokers can choose to operate on a split of commission fees earned with traditional

splits being 50/50, 60/40, and 70/30. If this compensation package is chosen by the broker, they

do not pay the firm a periodic fee and simply splits their earned commissions on the agreed upon

basis with the firm. As an alternative to the split compensation package, the broker can choose

to pay their firm a periodic fee, usually monthly, and keep all of the earned commissions.

Technically, in this work, all brokers that face periodic expenses regardless of commissions

earned are defined as 100% brokers. All brokers that do not face periodic expenses and earn a

split/percentage of their generated commissions are defined as SPLIT brokers. These

classifications allow this piece to capture for the first time the impact of intrafirm compensation

on property price and marketing time.
A residential real estate broker, generally speaking, receives a higher percentage of the

commission generated as their production increases. However, this work hypothesizes only

brokers with superior skills and ability will choose the 100% compensation package. Brokers

with relatively inferior skills and ability will opt for the less risky SPLIT compensation package.


A model is outlined below in which brokers and their firms negotiate next period’s compensation

plan. The commission structure offered by firms to its brokers generally falls along two lines. A

broker can contract with a firm as either a SPLIT or 100% broker. Given the firms ability to

costlessly recognize a particular broker’s production via MLS records, there is symmetry of

essential information between the negotiating parties over closed sales volume. That is to say,

the firm knows the broker’s production capabilities and the broker knows the firm has this

information. Brokers, however, have a conversion ratio, , known only to themselves.

Conversion ratios are the rate at which brokers convert prospective buyer and sellers into closed

properties and thereby earned commissions. The remaining critical assumptions are outlined


        1. The firm is risk neutral.

        2. Brokers possess an aversion to risk. Specifically, they fear the inability to recapture

           the 100% compensation package’s periodic fee, .

        3. Negotiation of employment contracts and transfers of licenses, when necessary, are

       4. Brokers are utility maximizers. Specifically, more of a good thing, such as

           compensation (C), is better.

       5. There exists an optimal mix of SPLIT and 100% brokers. Perhaps because firms seek

           reputational capital that is adversely affected by failed/expired listing.

       6. The contingent fee, , charged sellers to market their property is competitively

           determined and taken as given by the market.

As mentioned above, two different compensation packages are typically offered the broker. The

SPLIT compensation package allows the broker to receive a percentage of the gross

commissions earned from closings and he/she does not pay the firm a periodic fee. A 100%

broker’s compensation plan is slightly different. The 100% broker pays the firm a periodic fee

and receives 100% of the gross commissions earned at closing.

These two compensation plans provide the opportunity to represent a general compensation plan

for both SPLIT and 100% brokers. The general compensation plan can be algebraically

expressed as follows:

       C =  -                                             (1)

Where, C represents a broker’s compensation in terms of dollars with -∞ < C < ∞. The term 

(0    1) represents the percentage of the gross commission earned by the broker. Obviously,

1-K represents the firm’s share. The term  represents a broker’s periodic sold volume in dollar

terms with  > 0. The term  (0    1) represents the commission in percentage terms
charged the seller of a particular property by the broker. Finally, the term  ( > 0) represents

the periodic fee in dollar terms paid by the broker to the firm. Understanding that  = 1 for

100% brokers and  = 0 for SPLIT brokers, equation (1) can be generalized into two distinct

compensation plans.

       CSPLIT =                                                   (2)

       C100%er =  –                                               (3)

CSPLIT represents the compensation for the broker who chooses the split compensation package,

while C100%er represents the compensation for the broker who opts for the 100% compensation

package. Equation (2) is presented in Figure I in C x  (compensation by sold dollar volume

space). Obviously, from equation (2) and Figure I, as a SPLIT broker closes greater dollar

volumes, charges higher fees, and negotiates higher splits, their compensation increases.
    Figure I


Figure II depicts the compensation arrangement for a 100% broker, equation (3), in C x  space.

As can be seen, the 100%er has exchanged a periodic fee, , for 100% of the gross commissions

earned, K = 1. Also, Figure II reveals the minimum sold dollar volume, , for the contracting

period. A broker choosing the 100% compensation package and closing less than  does not

recapture their paid periodic fee and is out of business. Brokers will avoid this potential problem

at all costs. Those unable to generate  will opt for the SPLIT compensation arrangement with

their firm.

                                              Figure II


                                                                                     

The two compensation plans are superimposed on one another in Figure III. Figure III depicts a

visual break-even analysis (*) as well as other interesting results. A viewing of Figure III

indicates that an increases in  (the commission split between a SPLIT broker and the firm) and

 (the periodic fee paid to the firm by a 100% broker) leads to an increase in * (the closed

volume necessary for a broker to be indifferent between a split and 100% compensation plan).

                                              Figure III



                                                            *                       

The compensation for a SPLIT broker originates from the origin, because a SPLIT broker does

not pay a periodic fee, and extends Northeast in the first Cartesian quadrant. Its slope is

determined by the value of , i.e., the percentage of the gross commission retained by a SPLIT

broker. As a SPLIT broker produces greater amounts of closed volume, his/her compensation


The compensation for a 100% broker originates from a negative intercept of  along the y-axis,

because of the periodic fee charged the 100%er, and proceeds Northeast with a slope of one (1)

from the fourth to the first Cartesian quadrant. The two compensation plans are simple linear

equation and are destined to intersect in the first quadrant producing a simple break-even

analysis for brokers negotiating with their firms over their next period’s compensation plan.

Equating (2) and (3) and solving for  reveals the necessary closed volume, *, for both brokers

and firms to be indifferent between the two compensation plans.

       * =  / (1 - )                                      (4)

A broker producing  >  / (1 - ) units would prefer the 100% compensation plan. A broker

producing  <  / (1 - ) units would prefer the split compensation plan. However, firms

seeking perhaps to enhance their reputational capital are not interested in all brokers who

produce *. Specifically, they are interested in brokers with high conversion ratios or more skill,

but a broker’s conversion ratio is unobservable to the firm. Yet, it is possible, in equilibrium, for
a firm to pick a  > * to induce brokers with the higher conversion ratios to self-select into that

firms 100% program.

To see this, let  (0    1) represent the probability of a successfully sold listing, i.e., a

broker’s conversion ratio. Furthermore, assume this conversion ratio is the same across all

listings for a particular broker within a given time period. Now, the expected compensation for a

broker can be calculated across the two compensation arrangements. Equations (2) and (3) need

only to be restated including .

        CSPLIT =                                                    (5)

        C100%er =  –                                                (6)

* can now be recalculated to consider broker conversion ratios. Equating equations (5) and (6)

reveals this new *.

        * =  / (1 - )                                              (7)

Obviously, only those brokers expecting to have an end of year closed volume greater than *

will prefer the 100% compensation package. Equation (7) can be rearranged to reveal the

equilibrium condition, which makes brokers and firms indifferent between the two compensation

packages in terms of . Said another way, the firm can pick a  such that only brokers with a

sufficiently high conversion ratio will choose the 100% compensation package. Calling this *,

we can calculate this periodic fee:
       * = (1 - )                                                (8)

Given  is market determined,  is observable for each broker, and (1 – K), (and therefore K) is

determined by way of competitive equilibrium. Firms can choose a  marginally larger than *

that will encourage low quality brokers (lower ’s) to signal their lower skills and separate from

high quality brokers (higher ’s). Said another way, high skilled brokers seeking to maximum

their compensation will self-select and become 100% brokers while less skilled brokers see the

cost of mimicking as to expensive and will choose to become SPLIT brokers. Specifically,

brokers with  >  / (1 - ) will opt for 100% programs, and those with  <  / (1 - )

will choose to become SPLITs.

These two distinct pools of brokers provide a perfect opportunity to test the implicit assumption

of broker equality found in many prior works. If brokers’ abilities and skill levels are uniform as

many suspect, neither pool should impact a property’s price or marketing time when a control is

instituted for the presence of 100% brokers in a traditional hedonic pricing or duration model. If,

on the other hand, a positive and significant coefficient for 100%er is found in the hedonic

pricing model and/or a negative and significant coefficient in the duration (time on market)

model varying abilities and skill levels amongst brokers is suggested.

Data and Models

The original data set consists of all conventional residential closings (2,716), that occurred

during the calendar year 1998 in Montgomery, Alabama. The Montgomery Area Association of
Realtors (MAAR) Multiple Listing Service (MLS) is the principal source of data for this study.

Ancillary information is provided by the Montgomery County Tax Assessor’s Office. MAAR

provides the data on selling price, selling time, location, and most of the physical characteristics

of listed properties. Because the MLS listings do not contain any information on the age and

square footage of listed properties, this information is obtained from the County Tax Assessor’s


To insure a complete set of housing characteristics for each home sale, observations that do not

appear in both the MLS and tax assessor’s database are eliminated. Next, obvious data entry

errors from the MLS database such as negative time on the market, zero bedrooms or baths, etc.

are eliminated from the data set. This leaves 1,549 observations for the period in question.

Descriptive statistics for the entire sample and a legend for variable definitions are presented in

Exhibits I and II, respectively. The sample is further subdivided into two subsets. The

descriptive statistics for properties marketed by brokers who have opted for the 100%

compensation package are formally reported in Exhibit I. While Exhibit II reports the

descriptive statistics for those properties marketed by brokers who have chosen the split

compensation package earlier described. When convenient, throughout the remainder of this

work, 100% or100%ers will be used to represent brokers receiving 100% of the gross

commissions in exchange for accepting certain periodic operating expenses. In addition, when

convenient, SPLIT or SPLITs will be used to represent brokers who have opted for the split

compensation plan.
Comparing the two sub-samples for 100%ers (Exhibit III) and SPLITs (Exhibit IV), one statistic

immediately stands out. The average marketing time for properties (TOM) sold by 100%ers

appears, at least casually, significantly shorter than properties sold by SPLITs – 70.82 days vs.

93.51 days. The average sales price (SP), on the other hand, though higher for 100%ers, does

not seem dramatically higher -- $118,061 vs. $114,186.

                                         Exhibit I
                          Summary Statistics Sub-sample – 100%ers

                      Variable N    Mean     Median StDev
                      SP        588 118061 102000       55045
                      TOM       588    70.82       57    59.15
                      AGE       588   21.546       19  19.312
                      SQFT      588   1833.4    1716     536.1
                      BED       588   3.2755        3  0.6196
                      BATH      588   2.2245        2  0.6182
                      JD        588   0.5646        1  0.4962
                      LEE       588   0.2789        0  0.4488
                      LANIER    588   0.1446        0    0.352
                      CARVER    588   0.0119        0 0.10855
                      DRIVE     588   0.4609        0    0.519
                      GAR       588   0.3146        0  0.4648
                      CPT       588   0.2245        0  0.4176
                      FP        588   0.8265        1  0.3834
                      GTUB      588   0.3248        0  0.4687
                      SEPSHOW   588   0.2874        0  0.4529
                      POOL      588   0.1037        0  0.3052
                      DOUBOVN   588   0.1037        0  0.3052
                      EIFS      588 0.04592         0 0.20949
                      NOMKT     588   0.0646        0  0.2461
                      NC        588   0.0748        0  0.2633
               Exhibit II
 Summary Statistics Sub-sample – SPLITS

Variable N    Mean     Median StDev
SP        961 114186     98000   53345
TOM       961    93.51      75    75.12
AGE       961 20.454        19 17.292
SQFT      961 1815.1      1697    532.8
BED       961 3.2268         3 0.5688
BATH      961 2.1717         2 0.5833
JD        961 0.5005         1 0.5003
LEE       961 0.3611         0 0.4806
LANIER    961 0.1238         0 0.3296
CARVER    961 0.01457        0 0.11988
DRIVE     961 0.4828         0 0.5164
GAR       961 0.2758         0 0.4471
CPT       961 0.2414         0 0.4282
FP        961 0.7534         1 0.4337
GTUB      961 0.3236         0 0.4681
SEPSHOW   961 0.3007         0 0.4588
POOL      961 0.1103         0 0.3134
DOUBOVN 961      0.077       0 0.26674
EIFS      961 0.05411        0 0.22635
NOMKT     961 0.03642        0 0.18743
NC        961 0.04787        0 0.2136
These differences could be randomly generated and/or be due to the fact that 100% somehow

systematically list and sell properties that on average take less time to transact and that on

average sell for higher prices. In order to account for these eventualities, this piece develops

hedonic pricing and time-on-market models to control for these possibilities and other sources of

spurious variability. Exhibit III provides a Legend of variable names and descriptions and

Exhibits IV and V provide models of the impact of 100%ers on property price and marketing

time, respectively.

                                           Exhibit III
                                         Variable Legend

SP  contract sales price of the property
TOM  time on market in days
AGE  age of the property
SQFT  square footage of the property
BED  number of bedrooms in the property
BATH  number of baths in the property
JD  one if the property is in the Jefferson Davis school, zero otherwise
LEE  one if the property is in the Lee school zone, zero otherwise
LANIER  one if the property is in the Lanier school zone, zero otherwise
CARVER  one if the property is in the Carver school zone, zero otherwise
DRIVE  one if the property has a driveway only, zero otherwise
GAR  one if the property has a garage, zero otherwise
CPT  one if the property has a carport, zero otherwise
FP  one if the property has a fireplace, zero otherwise
GB  one if the property has a garden bath, zero otherwise
SEPSHOW  one if the property has a shower separate from the tub, zero otherwise
POOL  one if the property has a pool, zero otherwise
DOUBOVN  one if the property has a double oven, zero otherwise
EIFS  one if the property has EIFS, zero otherwise
NOMKT  one if the property was not traditionally marketed, zero otherwise
NC  one if the property is newly constructed, zero otherwise
SPLIT  one if the property was marketed by a SPLIT broker, zero otherwise
100%er  one if the property was marketed by a 100%er, zero otherwise
                                         Exhibit IV
                        Pricing Equation – Dependent Variable LnSP
                     Predictor    Coef     StDev T       P       VIF
                     Constant       8.7994 0.1725 51.020 0.0000
                     LnAGE         -0.0255 0.0069 -3.710 0.0000 1.5
                     LnSQFT         0.2504 0.0235 10.660 0.0000 1.3
                     LnBED          0.3302 0.0342 9.660 0.0000 1.3
                     LnBATH         0.4657 0.0258 18.030 0.0000 1.5
                     LEE           -0.0330 0.0132 -2.500 0.0130 1.2
                     LANIER        -0.1212 0.0190 -6.390 0.0000 1.3
                     CARVER        -0.1772 0.0497 -3.560 0.0000 1.1
                     GAR            0.1970 0.0144 13.670 0.0000 1.4
                     CPT            0.0874 0.0144 6.070 0.0000 1.2
                     FP             0.1297 0.0151 8.590 0.0000 1.3
                     GTUB           0.0646 0.0151 4.270 0.0000 1.6
                     SEPSHOW        0.1579 0.0160 9.880 0.0000 1.7
                     POOL           0.0970 0.0185 5.250 0.0000 1.1
                     DOUBOVN        0.0632 0.0207 3.050 0.0020 1.1
                     EIFS           0.1399 0.0272 5.140 0.0000 1.2
                     NOMKT          0.0371 0.0369 1.010 0.3140 2.0
                     100%er         0.0027 0.0118 0.230 0.8170 1.0
                     LnTOM         -0.0020 0.0060 -0.330 0.7410 1.9
                     R2              72.00
                     F-Stat         218.99                0.0000
                     N                1549

Exhibit IV indicates that 100% brokers, on average, sell property at a premium relative to SPLIT

agents, notice the positive and statistically significant coefficient on 100%er.
                                            Exhibit V
                            TOM Equation – Dependent Variable LnTOM
                        Predictor  Coef      StDev T        P       VIF
                        Constant     3.8600 1.2080    3.200 0.0010
                        LnAGE       -0.0259 0.0295 -0.880 0.3810 1.5
                        LnSQFT       0.0358 0.1041    0.340 0.7310 1.4
                        LnBED        0.2987 0.1503    1.990 0.0470 1.4
                        LnBATH       0.2660 0.1214    2.190 0.0290 1.8
                        LEE         -0.1458 0.0566 -2.580 0.0100 1.2
                        LANIER       0.0145 0.0821    0.180 0.8600 1.4
                        CARVER       0.3293 0.2134    1.540 0.1230 1.1
                        GAR          0.1095 0.0652    1.680 0.0930 1.5
                        CPT         -0.0845 0.0622 -1.360 0.1750 1.2
                        FP           0.0961 0.0661    1.450 0.1460 1.3
                        GTUB         0.0991 0.0650    1.520 0.1280 1.6
                        SEPSHOW     -0.2219 0.0703 -3.160 0.0020 1.8
                        POOL        -0.0218 0.0797 -0.270 0.7850 1.1
                        DOUBOVN -0.0683 0.0888 -0.770 0.4420 1.1
                        EIFS         0.3268 0.1172    2.790 0.0050 1.2
                        NOMKT       -4.1379 0.1171 -35.350 0.0000 1.1
                        100%er      -0.2521 0.0498 -5.060 0.0000 1.0
                        LnSP        -0.0361 0.1093    -0.33 0.7410 3.6
                        R               48.1
                        F-Stat        78.66                  0.0000
                        N              1549

Exhibit IV indicates that 100% brokers, on average, sell property quicker than SPLIT agents,

notice the negative and statistically significant coefficient on 100%er.

Thus, it is possible for real estate firms to pick a periodic fee (Δ) that causes the brokers with the

highest conversion ratios () to self-select into a 100% compensation package. Additionally,

there is convincing evidence that these 100% brokers can sell property faster and in less time on


 In this piece, the terms broker and agent are used interchangeably. This is done for ease of exposition. In practice,
however, these terms are used to differentiate between different levels of licensure.

Shared By: