Quant-in-R

Simple financial applications & analysis programs written in R language

SHY: The 'SHarpe Yield' Strategy // SHY strategy on the whole Taiwan Stock Market

R, as a programming language

The R programming language is designed for statistical analysis, and is widely used in many area recently. Here, I will focus on one of the most adopting domain, financial analysis, to demonstrate the strength of R. Besides, I shall pay extra attention to data visualizaiton in R.

The SHY 'SHarpe Yield' strategy

The introduced investment strategy, SHY 'SHarpe Yield', has a very simple computation model. It evaluates each stock by the SHY value, simply yield represents in a sharpe ratio way, i.e., its mean divided by its standard deviation. Given a list of sample stocks in folder csv/ (a dozen stock csv files reside), we can compute the SHY values and report the SHY value as follows:

    source("shy.R")
    shy_list = get_shy_suggestion()

Then, it shows

[1] "( 1) 個股：2382, ＳＨＹ：3.529709"
[1] "( 2) 個股：2912, ＳＨＹ：3.470507"
[1] "( 3) 個股：1101, ＳＨＹ：3.037783"
[1] "( 4) 個股：2330, ＳＨＹ：2.506207"
[1] "( 5) 個股：2207, ＳＨＹ：2.486131"
[1] "( 6) 個股：2801, ＳＨＹ：3.330069"
[1] "( 7) 個股：2701, ＳＨＹ：3.978262"
[1] "( 8) 個股：2357, ＳＨＹ：2.116582"
[1] "( 9) 個股：2325, ＳＨＹ：2.385228"
[1] "(10) 個股：1434, ＳＨＹ：3.511784"
[1] "(11) 個股：2354, ＳＨＹ：1.811173"
[1] "(12) 個股：0050, ＳＨＹ：NA"
[1] "(13) 個股：2408, ＳＨＹ：NA"

Based on the computed SHY value, we can form a strategy that selects stocks with better SHY, for example, buy a bunch of stocks with better SHY value periodically, just as the way as ETF (Exchange Traded Funds) does.

How good (or bad) is SHY?

Would you like to believe a strategy is good when someone throws it to you and claims that it is? I hope not.
Here, PerformanceAnalytics, a powerful module in finance evaluation in R is adopted, to express the quality of SHY:

    source("shy_eval.R")
    eval_res = evaluate_shy()

The result is then shown in your RStudio console:
(Note: here we buy 3 stocks based on SHY values periodically, among the 11 candidates)

In case that the performance comparison between two targets is interested, we can put the two targets in same evaluation:

    eval_pair_res = evaluate_shy_stock('2330')

Another way to compare performance is to use a relative performance view:

    eval_rel_res = evaluate_shy_stock_relative('2330')

If some stocks, instead of SHY are interested, we can do the following:

    source('stock_eval.R')
    evaluate_stock('2330')
    evaluate_stock_pair('2330', '2357')
    evaluate_stock_pair_relative('2330', '2357')

Don't forget to try above yourself to see how the result is. :)

Note: have 'DEBUG=1' to show more message during computation

Web Visualization

Visualization is always a good idea for better comprehension, especially when it is presented by a web page. Shiny is yet one perfect web visualization framework to help R programmers present their work as a web page, in less effort without downgraded outcome. A page on how SHY performs, under the whole Taiwan Stock Market, can be found at:

SHY: The 'SHarpe Yield' Strategy // SHY strategy on the whole Taiwan Stock Market

There is also a page for simple data visualization, which you can open it locally as follows:

    library(shiny)
    setwd("SHY_draft")
    runApp()

Or visit the online page: SHY_draft: Simple demonstration for data visualization

More R programs

For future plan, I will present programs utilize some other R features in terms of financial analysis:

• Probability Distributions
• Linear Least Square Regression
• Confidence Intervals
• pValues
• Power of a test