Nirvana User Guide

Introduction

This 90 seconds video is a very quick demonstration of most features of the portal.
Hopefully this motivates you yo explore further.

Objective

Before getting in the nitty-gritties of things, I would just like to take a moment to explain the purpose of developing these tools in the first place. As a retail investor I always found myself severely limited by the availability of research on major financial portals and their experts. Gathering original insights in quick and easy way seemed like something of entitlement to me that was unfulfilled. That led me to explore what raw resources are available and price action data was the most obvious one. So that's where we start.

After experimenting with some data recreationally some interesting concepts emerged. And then some more. Seemed to get better with every iteration.

Instead of going through the numbers crunched by my scripts, I found making and analysing charts much more interesting and efficient. By making these charts 'live', their powers increased ten folds. By opening up the tools for public consumption, I would like to make the iterations much faster and much better while helping like minded people at the same time.

This is just a tiny first step in what really is extremely ambitious vision of visual analytics for the global capital markets.

Introduction

Interactivity of Nirvana could come as a surprise to many who have seen charts only in ppts and may be as static pictures on webpages. Clicking on a chart, hovering to see the details, dragging to zoom-in and out never occurs to most.

But the interactivity is the heart and soul of Nirvana. Without it it's just a bunch of charts with seemingly no purpose. This guide explains various views, charts and their purpose. Case studies section will shed light on how this portal can be used for drawing conclusions for specific research cases.

Please note, at the moment this portal is limited 500 stocks in NSE500 index. But it is immensely scalable and will be adding new markets and instruments in time.

Views Basics

At the moment the app is divided into three basic sections or view. Indices, Equities and Quick Stats. Here is basic introduction. We will get into details in further sections.

Indices

Indices view deals with NSE broad market (Nifty50) and sector (e.g. CNX_METALS) indices. This is meant to serve as a quick view of where the markets stand and their journey in time.

Equities

Equities view starts with equities and they way they are categorized in industries. To begin with one sees a chart showing all 500 stocks in a single plot! And this chart is of course interactive.

Quick Stats

While the indices and equities views are designed to facilitate forming one's own conclusions, the quick stats view is for instant gratification. This view lists stocks satisfying certain criteria without having to go through all the charts. These are meant for immediate consumption and may be identify candidates for further research and make quick investment decisions.

Indices View

NSE sector indices are a great view to dissect various segment of markets and shortlist for further analyses. In this view, we have some plots of the sector indices to begin with and then ways to drill down to specific stocks for zooming in on investment opportunities.

The first plot is a simple line chart of sector indices over past few months. It shows on two indices on load to avoid clutter, but it has 9 indices in all. One can select/deselect the indices by clicking on the index name at the bottom. To select all the indices or a specific one double click on the index name.

To make the price action comparable, indices are rebased to 100 on start date. The start date and the end date both can be customised on datepicker fields on top. This helps in analysing sector performance over various periods of time. Following gif demonstrates the functionality.

Sector Indices

Returns

The next plot shows returns of sector indices over previous quarter, month and day all in one simple and intuitive chart. This is useful not only to quickly see how the performance has been over recent past but also to identify beginning or ending of trends. For example when a particular sector shows very good quarterly returns but the monthly and daily returns are not not contributing over a few days, it's a fair guess the rally is tiring and consolidation or correction may be in order.

Clicking on segments of this bar chart opens up the index and shows stocks belonging to that index for further analyses in the chart below. e.g. Click on quarterly segment of the Realty index in bar chart automatically selects the Realty index and quarterly period in the chart below. We will cover this in more details later.

The following gif demonstrates the returns plot.

Index Returns

Index Stocks

This plot and its variants are used quite a bit in this app. It shows all the stocks of a particular index, their volatility over a given period plotted against returns over the period.

Both the index and periodicity can be customised from dropdown and radio buttons on top.

The plot also shows a linear regression line and basic regression output. When there is a strong trend exhibited by the stocks of an index, slope would be high positive/negative number depending trend and R-square would be close to 1. When there is no discernible trend in the sector as a whole the regression line would be flat and R-square close to 0.

Each point on the plot represents a stock and hovering above the points reveal the ticker name, industry the stock belongs to and also returns and vol of the stock for the particular period.

Clicking on a point opens up detailed view for that particular stock below. This again is fairly reused across the app. We will cover the in more detail separately. In short, it shows chart of the stock, key price action stats and distribution of returns over various periods.

Index Stocks

Indices' Correlations

The portfolio theory suggests to have a diversified portfolio, one needs to invest in stocks with low correlations. We are extending this to indices as it is reasonable to assume stocks within index would have roughly similar characteristics.

This plot shows correlations of indices against others. The stacked bar plot has 9 columns representing each index. Each column in turn shows correlation against each other index (and also itself). The periodicity is adjustable using radio buttons above.

This is very useful in identifying market regimes in general. In a raging bull market everything will go up. In a panic stricken market everything will go down. In both these cases correlations would be high and height of the bars would be high and almost even.

In a stable market, correlations would show some diversification and we would see some boxes below the zero line implying negative correlations for some sectors.

Interesting insights could be gained by just observing this chart over a period and seeing it change. Also, divergence between quarterly and monthly plots could signal shift in market conditions.

Weekly chart is useful for short term trading and also analysing how markets behave around key events (like option expiry or RBI policy statement).

Hovering on the chart shows correlations of those indices. Clicking on them would open a chart below that plots stocks of both the indices together with different colors for quick visual cues and further analyses.

Indices Correlations

Equities View

As the name suggests this view focuses on individual equities first and then tries to discern any hypotheses based on price action of individual securities.

This is opposite of what we did in the indices view. It was more of a top down approach whereas this one is bottom up.

All Stocks

True to the point, the opening plot shows all the 500 stocks of NSE500 index in one plot. Nice, eh!

This one again plots stocks based on their returns against their volatility for specific periods. The default is monthly returns against monthly vol. But one can change this to quarterly or daily by selecting radio buttons on top.

Note: Daily returns are plotted against monthly vols.

In the rawest form, this shows the returns at the price of risk for the entire market in one go. But the real fun is in zooming in and out of the chart by simple drags and finding stocks belonging to particular spectrum of risks and returns. Various investment strategies require stocks with different characteristics across this spectrum and identifying candidates for these strategies is a breeze using this simple plot.

For the sake of simplicity, let's consider a simple example. Say, one is interested in taking leveraged position using futures or may be just using margin in cash markets. Ideally, one would need to limit the vol for this kind of strategy. So what we really want is a few candidates with relatively low vol and highest (lowest) returns to go long (short). To do this we just need to focus on left most portion of the chart and select a few candidates for further research.

Again as we did in the indices view, hovering over the stocks reveals key points about the stock. Clicking on a point would open a detailed stock view below the plot.

As useful as this plot is for quick identification of trades, it is also fun just to randomly explore without any set objective. Try this on a idle Friday afternoon. Let the market speak to you!

All Stocks

Broad Market Stats

From the individual stocks now we move on to broader market. While deeply researching individual stocks is rewarding and necessary, it is crucial that one sees the big picture first.

Identifying market conditions by looking at market as a whole helps us identify market regimes (risk on/off, too much complacency or panic etc.) and avoid stupid investment decisions at wrong time.

These charts may look too simplistic but it's important to remember we are taking a bird's eye view on this. Most times, these will reveal nothing. And that's a great thing! We are looking for that odd behavior and odd times and we need to see a red flag when it happens.

The first thing to look at is distributions of returns of individual securities in markets for different periodicities. The chart shows quarterly, monthly and daily returns histograms plotted together. In a stable market shapes of individual distributions should be fairly consistent. We can observe that by selecting/deselecting periodicities from the chart. What we are looking for is any shifts in skewness (i.e. whether the distributions are leaning discernably left or right). Heights (kurtosis) are not too relevant if the shape is reasonably symmetric.

When regimes change, we observe some skewness in daily distributions consistently over some period. When that begins reflecting in monthly distributions and there is some divergence in quarterly and monthly skews, it's time to revisit your investment thesis!

Next three stacked bar charts are a play on similar themes.

Out of the 500 stocks in our universe, observing the market breadth over a period can give clues when status quo begins to change. At the moment of writing, markets are stable and we can see that in the gif below. There are about even number of stocks with positive and negative daily returns over past 20 days. Height of positive and negative bars are about even. Moving average of the positive and negative stocks are also evenly places on both sides of zero line. This is a calm market which is great for trading but not for the demonstration of this plot :) When things go bad or there is sudden positive mood in the market, we would see the bars crossing above or below the moving average lines and bias would be very obvious in the heights of bars contrasting recent history.

So even if one has not really kept abreast of markets in some time, this would still convey the picture as it exists.

The other two charts are similar. Instead of number of positive/negative returns, they convey number of stocks making monthly highs/lows and number of stocks with returns above/below 2 standard deviations of their average returns over a month.

While the bar charts are similar, sometimes they can tell a different story or give a prescient warning or a heads up. I do not want to give specific examples at this point because what one draws from this depends on their perspective. A bit like tarot card reading if you will.

Broad Markets

Stocks By Industry

We categorise stocks based on their industry so we could draw some inferences about industrywide performances. While this may be similar to sector indices, industry has a lot more participation. Sector indices have a limited number of stocks and as such are represented by few big names. Every stock in NSE500 has industry classification on the other hand and each industry is represented by a lot many stocks.

The first plot is similar to the positive/negative stocks of previous section. But this is just for the day. On this particular day it's looking rather uninteresting as it's calm and relatively flat market. But on eventful days this one has a story to tell. It shows number of stocks in positive/negative and monthly volatility of the industry.

Note: Volatility in this context is average volatility of all stocks belonging to that particular industry. This may not go down well with purists. Our objective is to get some sense of the vol rather than accurate estimate of vol of equal weighted portfolio. This context is important.

The second plot is similar to Returns section in Indices View. We have quarterly, monthly and daily average returns for the industry. This one behaves exactly in the same fashion. On hover it shows key information and on click it would open up the industry in the chart below and plot all the stocks belonging to the industry. The return period in the chart is the same as the one clicked.

Industry stocks plot should look familiar. It behave exactly like indices stock plot.

Following demo explains the usage.

Stocks By Industry

Quick Stats

While this app exists to help people with some interest in capital markets do their own research quickly and effortlessly, we understand not everyone is all that interested in that kind of stuff.

If you want all the gains without any pain, we've got your back too.

This section requires practically no explanation. We have laid out a bunch of tables listing top 5 stocks for specified categories. All you need to do is click on the stock name and see how the price action matches up and other characteristics of the stock.