/
munitrades.Rpres
90 lines (67 loc) · 3.36 KB
/
munitrades.Rpres
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
munitrades
========================================================
author: Marc Joffe
date: April 2, 2014
Background
========================================================
- Municipal bonds are issued by state and local governments to finance infrastructure.
- Typically, municipal bonds pay periodic interest. The annual interest rate is called the coupon.
- On the maturity date, the face value of the municipal bond becomes due.
- Before maturity, the bonds can be traded. Trade prices are expressed as a percentage of face value multiplied by 100.
- If a bond is trading at face value - or par - its price is 100.
The Data
========================================================
- A professor gave me a SAS file containing 97 million trade records including:
- Bond information: Bond identifiers, Maturity dates, Coupon rates
- Trade information: Trade size, Trade price
- I converted the SAS file (with extension SAS7DB) to delimited text and imported to PostrgeSQL
- I found that the vast majority of records were duplicates,
- A SELECT DISTINCT query in PostgreSQL, revealed only 1.1 million unique rows
- I imported the unique rows to R using the package RPostgreSQL
Loading the Data
========================================================
```{r, eval=FALSE}
library(RPostgreSQL)
drv <- dbDriver('PostgreSQL')
con <- dbConnect(drv, dbname='msrb', user='postgres', password='********')
munitrades <-dbReadTable(con,'munitrades_unique')
```
Price Distribution
========================================================
Nearly bell shaped curve with a mean of 97.58
Determinants of Price
========================================================
```{r, eval=FALSE}
# Calculate the days remaining before maturity as of the trade date
munitrades$remaining_term <- as.numeric(munitrades$maturity_dt) - as.numeric(munitrades$tradedate_dt)
regr <- lm(price_inter ~ coupon + remaining_term, data=munitrades, na.action=na.exclude)
summary(regr)
```
<pre>
+----------------+-------------+-------------+---------+-----------+
| Coefficients | Estimate | Std. Error | t value | Pr(>|t|) |
+----------------+-------------+-------------+---------+-----------+
| (Intercept) | 88.79 | 0.03132 | 2835.3 | <2e-16 |
| coupon | 2.365 | 0.005088 | 464.8 | <2e-16 |
| remaining_term | -0.0005129 | 0.000002334 | -219.7 | <2e-16 |
+----------------+-------------+-------------+---------+-----------+
Adjusted R-squared: 0.1742
</pre>
After **regr** runs, the **fitted** vector contains modeled prices
Are Small Investors Getting a Bad Deal?
========================================================
- Strategy
- Find all securities which had at least 10 trades
- Break up the trades into two groups:
- Small trades (below the median trade size for that security)
- Large trades (above the median)
- Compare the variances of the trade prices for the small and large groups
- dplyR library facilitates this type of analysis by providing SQL and Excel pivot-table functionality in R
Results
========================================================
Also: Paired t-test concludes with 95% confidence that the variances of the larger trades are lower than those of the smaller trades
```{r, eval=FALSE}
t.test(casted$abovemedian,casted$belowmedian,paired=TRUE)
```