As far as investors and traders are concerned, elections fall into the category of known unknowns. You know that an election will affect investment patterns and hence, lead to price volatility. But you don't know the outcome of the election and you don't know in what direction the volatility will go.
Nate Silver, the brilliant American psephologist, titled his best seller, “The Signal and the Noise”-and indeed, predicting who is likely to win an election is very much about trying to discern a coherent signal, or pattern in the middle of what seems like a lot of noise. It is very easy to see spurious patterns when looking at masses of data and it is also easy to miss a valid but weak pattern.
Americans like Silver have it relatively easy when it comes to predicting election results. There are only two major parties-the Republican and Democrat. So, even a random choice of winner has a 50 per cent shot at getting it right.
The vote-shares of those two parties don't change a great deal. It is even more helpful that both American parties have very democratic internal organisations, electing the candidates for all sorts of tickets. Any US citizen can sign up and become a member of either party. That citizen may stand for internal party elections, if he or she aspires to any electoral ticket, ranging from local municipal councillor to President.
This internal democratic process generates a lot of useful data. The psephologist knows how many party members there are in a given state, how much money a given candidate has raised, who is campaigning on what policy platforms, etc. Even under these conditions, pundits often get US election predictions wrong.
India has a much more difficult to predict, multi-party, first past the post electoral system. No Indian party is democratic in terms of internal processes. Candidates are selected by the High Command or the party boss, whoever he or she may be. It is impossible to get hard data on how much money a given Indian candidate can raise.
In mathematical terms, predicting a race with multiple candidates is many orders more difficult. Indian opinion polls are also conducted in opaque ways. Pollsters don't reveal sampling patterns, the questions asked and their order, the assumptions made in converting vote-share to number seats, and other crucial details. To add to the problems, Indian elections often see seats being won by extremely narrow margins, which are less than the inherent error margins of any opinion poll.
Given all those factors, betting on Indian election outcome with any high degree of confidence, is extremely difficult. However, since we know that a general election is likely to generate high volatility across markets, it is worth trying to find out how markets respond to this political process.
Making accurate predictions about the 2014 results is difficult. But we can extrapolate some things. One is that it's clear that the market would like a BJP-led government with Narendra Modi in charge. If it does get this, the current bull run is likely to be extended and to strengthen for a while at least. If it doesn't get this, the market could see a correction.
History helps us make some other projections. India has not seen a single-party majority since 1984. It is extremely unlikely that this pattern will be broken. The NDA may get a stable government with only a few BJP allies propping up Modi. It might get a more fragile alliance with multiple parties, in which case, Modi's position would be somewhat weakened. Or, there could be a khichdi Third Front backed by the Congress or the Congress may be backed by a bunch of non-BJP parties. The markets want either the first or second case.
We can take a look at market movements going into the last five general elections. I chose to look at the Nifty, starting from a month before polling commenced, running through the month of polling and carrying on through the “honeymoon” period of the first 100 days of the new Lok Sabha. This is a period of about 5.5 months-roughly 110-odd trading sessions. Outside this period, markets seem to revert to less sentimental behaviour, moving more on the basis of fundamentals. Take a look at the table below
How the Market reacts to elections
| Date | Returns until polling ended (%) | First 100 Days of new Lok Sabha (%) |
| 1996 | 9.88 | 3.44 |
| 1998 | 12.45 | -3.58 |
| 1999 | -0.53 | 9.61 |
| 2004 | 0.72 | -10.33 |
| 2009 | 18.79 | 52.34 |
| Returns computed from a month before polling started | ||
There is an apparent pattern of optimism about pre-poll trading, until the end of the last phase of voting. The returns were positive in most cases, the exception being the Kargil Election of 1999 and the “India Shining” election of 2004. The pattern is much more mixed if one considers the entire five-month period until the end of the honeymoon. There was one substantial gain, one substantial loss and one extraordinary gain.
However in 1999, the returns were partially driven by the global IT and Internet bubble of the time. In 2009, the incredible returns came as the global economy bounced due to easy money policies by various central banks after the subprime crisis.
The charts of Nifty movements during those five elections, normalised with the Nifty set to 100 a month before the elections can be seen in the graphs in the left.
Where is election 2014 likely to take the market?
The market zoomed to record levels and it has since consolidated. If the earlier pattern holds, we should see more gains going into the next six weeks before the voting ends. After that however, stock market returns over the 100-day honeymoon will probably depend on the results. An aggressive trader should be prepared to double up if the BJP takes enough seats to form a stable government. He should also be prepared to reserve direction and short (“double-minus” ) if a Third Front or UPA formation comes to centrestage.
The writer is an independent financial analyst.
This article was originally published on May 30, 2014.
Ask Value Research 
New episodes every Friday 
New Process Alert 
Seems like you've uploaded a wrong QR code. Please upload the QR code downloaded from MF Central platform.
