ShinyStocks.Rpres

Shiny Stocks
========================================================
author: Ruslan Kharitonov
date: January 24, 2015

Overview
========================================================

This Shiny Application helps to visualize Stock price 
and trading volume using 'quantmod' package.

- 'quntmod' package
- Deployment of ShinyStocks
- Visualizing 'AAPL' stock 
- Analyzing 'AAPL' stock
- Determining 'APPL' price direction 
- Conclusion

'quntmod' package
========================================================
- It is 'A rapid prototyping environment, 
where quant traders can quickly 
and cleanly explore and build trading models.'
- More details on http://www.quantmod.com/
- Used 'quntmod' in this project to simplify downloading and visualizing stock data.
- Installation of 'quntmod' package
```{r, eval=FALSE}
install.packages('quntmod')
```

Deployment of ShinyStocks
========================================================
- Could not use shyniapps.io due to it's lack of support for packages on 'R-Forge' repository
- 'zoo' is a dependency of 'quantmod', and it is on 'R-Forge' repository
- Created and configured Shiny Server on Amazon AWS EC2 Linux instance
- Used 'Ubuntu With R and Shiny (ami-5fe69f6f)' AMI to simplify the setup
- http://ec2-54-149-56-170.us-west-2.compute.amazonaws.com:3838/ShinyStocks/

Visualizing 'AAPL' Stock 
========================================================
- The Shine Stocks can be used to visual and analyze stocks
- Change 'Date range' will be reactive - recalculate the chart and other measures
- Example chart shows volatility of 'APPL' over 30 days

```{r, echo=FALSE}
library('quantmod')
data <- getSymbols('AAPL', 
                  from = '2014-11-24',
                  to = '2015-01-24',
                  auto.assign=FALSE)

   chartSeries(data, 
           theme = chartTheme("white"),
           TA = NULL
    )
```

Determining 'APPL' price direction  
========================================================
- Determining price by comparing average (mean) 
to the latest price ($112.98 per share).
- Will use the Adjusted close daily price (Ad)
- Price Direction Hypothesis: If latest price minus Adjusted Mean of a period is positive then direction is **Up**, otherwise it is **Down**.
   - ('2015-01-22' to '2015-01-24') Adj Mean: $112.69 **Up**
   - ('2015-01-19' to '2015-01-24') Adj Mean: $110.91 **Up**
   - ('2014-11-24' to '2015-01-24') Adj Mean: $111.86 **Up**
   - ('2014-01-24' to '2015-01-24') Adj Mean: $93.47 **Up**
   - ('2010-01-24' to '2015-01-24') Adj Mean: $64.79 **Up**

Conclusion 
========================================================
- chartSeries() can only be used with a period of 3 or more days, otherwise it produced an error.
- Need to take into account Stock splits (Adjusted Price) to determine correct average
- Latest price minus Adjusted Mean seems to predict historic direction of the stock price movement for 'APPL'
- It is unclear how the size of the period impacts the historic prediction
- Moving Average Convergence Divergence (MACD) is probably a better indicator of direction of a Stock price.
   - Detail on MACD: http://en.wikipedia.org/wiki/MACD 
- 'Past performance is not an indicator of future success'