Controlling
Individual Types
A big advantage of using
password files is that you lock each trader’s number in the market. For
example on the default password file the first five passwords
are as follows:
aa3247 |
ADD6790 |
adn3275 |
aj9134 |
AT5678 |
So the trader who logs in with aa3247 is trader number 1, ADD6790 is trader
number 2 and so on. It does not matter
when they login their slot is preserved.
This in turn lets you control precisely each trader’s type.
For example in the basic case
B01 there are two types defined as follows:
Trader
Type Data |
Type 1 |
Type 2 |
Cash |
1000 |
1000 |
Endow 1 |
15 |
0 |
Endow 2 |
0 |
15 |
Rights 1 |
0 |
0 |
Rights 2 |
0 |
0 |
Lower
Bound |
-100000000 |
-100000000 |
Lower
Bound Payoff |
-100000000 |
-100000000 |
Upper
Bound |
100000000 |
100000000 |
Upper Bound
Payoff |
100000000 |
100000000 |
Constant |
0 |
0 |
Linear
Coefficient |
0.0001 |
0.0001 |
Quadratic
Coefficient |
0 |
0 |
Log
Coefficient |
0 |
0 |
CARA
Coefficient |
0 |
0 |
CARA
Exponent |
0 |
0 |
HARA
Coefficient |
0 |
0 |
HARA
Exponent |
0 |
0 |
Each trader type has 1000
cash and trader type 1 has 15 units of endow 1 (coupon bond) and 0 units of the
zero coupon bond. Trader type 2 has 0
units of the coupon bond and 15 units of the zero coupon bond. Their trading rights are 0 for each (no
restrictions).
Aside: Trading rights can be set per security and
per trader: 0 no restrictions, 1 make market only (i.e., a dealer), 2 take market only, 3
cannot trade at all.
A linear conversion is used
for both types to trading or grade cash bonus.
This is merely 0.0001*market cash of their final marked position. Some interesting and simple variations of
this is to use a natural logarithm coefficient (say 1000 beside the log
coefficient) this will scale market cash as 1000*ln(Market Cash) and setting
it to 0 if market cash is negative. The
advantage of this is that it makes the traders attend to risk and return as if
they are risk averse with log preferences.
In the above the upper and
lower bounds define what range of market cash is permitted. This can be large. Outside the range the trading bonus (Grade
cash) conversion is bounded at the amount that corresponds to the upper/lower
bound.
Locking Types:
In the above the first five
passwords lock in the types as follows:
Aa3247 = Type 1,
ADD6790 = Type 2,
Adn3275 = Type 1,
Aj9134 = Type 2, etc.,
This is assigned irrespective
of when they actually login or what order they login.
So you can see how to control
the types and or type of market (dealer versus double auction) by specifying
each type carefully.
How Do I Add Types?
You can easily add new types
to any case. To do so requires
using additional columns carefully in the spreadsheet as illustrated below:
First, Cell B6 specifies how
many types. The default for B01 is 2
Second, A,B,C
20 --- A,B,C 47 must exactly equal the two types plus the labels.
This is illustrated below:
Number of
Securities |
2 |
Name |
Cp Bnd |
Maximum
Number of Trials |
8 |
Security
Type |
Bond |
Number of
Periods per Trial |
3 |
Price
Quotes |
Endogenous |
Period
Length (seconds) |
240 |
Start Life |
1 |
Maximum
Number of Traders |
60 |
End Life |
3 |
Number of
Trader Types |
2 |
Short
selling |
Yes |
Market
Depth |
10 |
Quote to
Price Formula |
Quote |
Depth
Displayed |
5 |
Information |
No |
Borrowing
Allowed |
Yes |
|
|
Last Row
with Exogenous Prices |
|
|
|
Number of
Information Types |
0 |
|
|
Last Row
with Information |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Recalculate |
No |
|
|
|
|
|
|
|
|
|
|
Trader
Type Data |
Type 1 |
Type 2 |
Payoff and
Settlement Data |
Cash |
1000 |
1000 |
Trial 1 |
Endow 1 |
15 |
0 |
Period 2 |
Endow 2 |
0 |
15 |
Period 3 |
Rights 1 |
0 |
0 |
Trial 2 |
Rights 2 |
0 |
0 |
Period 2 |
Lower
Bound |
-100000000 |
-100000000 |
Period 3 |
Lower
Bound Payoff |
-100000000 |
-100000000 |
Trial 3 |
Upper
Bound |
100000000 |
100000000 |
Period 2 |
Upper
Bound Payoff |
100000000 |
100000000 |
Period 3 |
Constant |
0 |
0 |
Trial 4 |
Linear
Coefficient |
0.0001 |
0.0001 |
Period 2 |
Quadratic
Coefficient |
0 |
0 |
Period 3 |
Log
Coefficient |
0 |
0 |
Trial 5 |
CARA
Coefficient |
0 |
0 |
Period 2 |
CARA
Exponent |
0 |
0 |
Period 3 |
HARA
Coefficient |
0 |
0 |
Trial 6 |
HARA
Exponent |
0 |
0 |
Period 2 |
Now suppose you want to
change from 2 types to 4 types, you do so as follows:
First cut the tan section
that is beside the types above and move this subsection over three
columns. Then copy and paste the types
section (two traders only not names) and paste beside type 2. Relabel types to 3
and 4 and change as desired the components.
Finally, do not forget to
change cell B6 from 2 to 4. This will
now appear as follws (plus we have given 3 and 4 a real position 1,000,000 cash and 15,000 securities!
Number of
Securities |
2 |
Name |
Cp Bnd |
Zero Cp |
|
Maximum
Number of Trials |
8 |
Security
Type |
Bond |
Bond |
|
Number of
Periods per Trial |
3 |
Price
Quotes |
Endogenous |
Endogenous |
|
Period
Length (seconds) |
240 |
Start Life |
1 |
1 |
|
Maximum
Number of Traders |
60 |
End Life |
3 |
3 |
|
Number of
Trader Types |
4 |
Short
selling |
Yes |
Yes |
|
Market
Depth |
10 |
Quote to
Price Formula |
Quote |
Quote |
|
Depth
Displayed |
5 |
Information |
No |
No |
|
Borrowing
Allowed |
Yes |
|
|
|
|
Last Row
with Exogenous Prices |
|
|
|
|
|
Number of
Information Types |
0 |
|
|
|
|
Last Row
with Information |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Recalculate |
No |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Trader
Type Data |
Type 1 |
Type 2 |
Type 3 |
Type 4 |
Payoff and
Settlement Data |
Cash |
1000 |
1000 |
1000000 |
1000000 |
Trial 1 |
Endow 1 |
15 |
0 |
15000 |
0 |
Period 2 |
Endow 2 |
0 |
15 |
0 |
15000 |
Period 3 |
Rights 1 |
0 |
0 |
0 |
0 |
Trial 2 |
Rights 2 |
0 |
0 |
0 |
0 |
Period 2 |
Lower
Bound |
-100000000 |
-100000000 |
-100000000 |
-100000000 |
Period 3 |
Lower
Bound Payoff |
-100000000 |
-100000000 |
-100000000 |
-100000000 |
Trial 3 |
Upper
Bound |
100000000 |
100000000 |
100000000 |
100000000 |
Period 2 |
Upper
Bound Payoff |
100000000 |
100000000 |
100000000 |
100000000 |
Period 3 |
Constant |
0 |
0 |
0 |
0 |
Trial 4 |
Linear
Coefficient |
0.0001 |
0.0001 |
0.0001 |
0.0001 |
Period 2 |
Quadratic
Coefficient |
0 |
0 |
0 |
0 |
Period 3 |
Log
Coefficient |
0 |
0 |
0 |
0 |
Trial 5 |
CARA
Coefficient |
0 |
0 |
0 |
0 |
Period 2 |
CARA
Exponent |
0 |
0 |
0 |
0 |
Period 3 |
HARA
Coefficient |
0 |
0 |
0 |
0 |
Trial 6 |
HARA
Exponent |
0 |
0 |
0 |
0 |
Period 2 |
Finally, whenever you change
a case save the Excel spreadsheet and run a local test on the same PC (i.e.,
open the market and two traders just to try it out).