FREE

India Strategy: Exit polls 2019 - NDA dominates across exit polls

Much awaited exit poll numbers are released and barring one, all of them are predicting a clear majority for NDA. Interestingly, few are predicting a clear majority for BJP (standalone) as well. We analyzed recent prediction accuracy across pollsters and shortlisted three polls with relatively best prediction rate (Exhibit 3-8). These three polls indicate an average 330 and a median 350 seats for NDA, indicating a possible full standalone majority for BJP. We saw a wide variance in estimates across key states of West Bengal, Uttar Pradesh and Tamil Nadu. Amongst the three shortlisted agencies (with best prediction rate), C-voter has most conservative estimates for these three states (mentioned above) and predicts and national tally of 287 for NDA. Also, C-voter indicates a possible gain of 75 seats for UPA and about 36 seats for NDA in post poll alliance. Interestingly, even if assume best case scenario for UPA from different polls, we don’t expect UPA to be able to form government, even with a gain of 75 seats by means of post poll alliance, as it would still fall short of majority. Hence, in most scenarios, we expect NDA to come to power with Mr. Modi at helm. Only vulnerability that we see arising is that NDA falls short of 272 mark and forms post poll alliance with other regional parties (which we expect to be stable). In that case the centralized decision making of the central government could take a back seat. We expect markets to read into exit poll results positively. While we believe global news flow (pertaining to US China trade war and geo-political stress in middle east) would pose bigger challenge to equity markets in near term, however, we still expect India to be in beneficial position as we expect India to benefit from dovish stance from global central bank (details ), especially as political uncertainty now gets done with. However, if outcome on 23^rd of May is different from exit polls summary (especially if NDA scores a lower number), we would brace for a negative reaction from equity and bond markets alike.

Best performing pollsters in recent past: In light of wide variance of predicted outcome by pollsters, we did a quick analysis on accuracy of prediction across pollsters across recent state elections and 2014 general elections. Our analysis indicates a good prediction rate by My Axis-India today, C-voter and Today’s Chanakya (Exhibit 3-8). Average of these agencies indicates 330 seats for NDA, while median indicates 350 seats (Exhibit 1) for the National Democratic Alliance (NDA). This average would mean a clear majority for BJP standalone as well.

Expect markets to be in high spirits, downside likely if outcome on 23^rd May is different: We expect markets to read into these numbers positively. While we believe there would be bigger challenge to markets from global news flows (pertaining to US China trade war and geo-political stress in middle-east), we expect India to be in beneficial position in case of positive capital flows (led by expected dovish stance by central banks). This would hold strongly for Indian economy now as political uncertainty gets done with. However, if outcome on 23^rd of May is different from exit polls summary (especially with NDA scoring lower number of seats), we would brace for a negative reaction from equity and bond markets alike. 

Wide variance across few key states, C-voter numbers most conservative amongst three agencies, indicating a floor, according to us: We had a quick look at the biggest swing states and their estimates by the above mentioned agencies. We observed the biggest variance in estimates for West Bengal, UP, Bihar and Tamil Nadu, with most C-voter having most conservative estimates across said states (Exhibit 2). Quick comment on key states/regions below:

• West Bengal: BJP seems to have established major inroads in the state. While, in 2014, it managed to bag 2 seats in the state, for 2019 estimates vary widely from 10-26, in either case indicating a strengthening hold of the party.
• UP and Bihar: UP and Bihar together account for 120 seats in Lok Sabha and are driven strongly by caste and religion based dynamics. UP is presenting the widest estimates ranging from 38-73, while Bihar exit polls are indicating strength for NDA across polls.
• Hindi Heartland: Non UP-Bihar Hindi heartland, comprising of Rajasthan, Madhya Pradesh, Gujarat and Chhattisgarh, seems to continue as strong bastion for BJP. While BJP won 88 seats in these four states in 2014, current estimate range varies from 74 to 94, indicating an ongoing strong hold for the Saffron party.
• Tamil Nadu: Tamil Nadu is an interesting state, where NDA is expected to lose maximum ground as AIADMK is expected to cede to DMK. While AIADMK win expectations vary from 0-10, that for DMK vary from 27-35. It’s interesting to note the emergence of other regional parties like MNM, which C-voter expects to grab as many as 11 seats.

Vulnerabilities and scope of going wrong: While most exit polls indicate ongoing strength for NDA, we believe the negative surprise here would mean NDA having maximum seats, but falling short of full majority. In that case we still expect NDA to form a government with post poll alliance. However, a coalition government (relatively low probability event) would mean Mr. Modi would still remain at helm but centralized decision making may take a back seat.

Possible post poll alliances for NDA and UPA: We also look at possible post poll alliance partners for UPA and NDA to understand the scenario in case 23^rd May outcome is drastically different from what is being predicted by exit polls.

• Possible NDA post poll alliance partners could be YSR, TRS, BJD and MNF and may add about 36 seats to NDA tally (as per C-voter) in case alliance works out.
• Possible UPA post poll alliance partners could be AIUDF, LDF, Mahagathbandhan (MGB) and TMC and may add as much as 75 seats to the tally in best case scenario (as per C-Voter). In such case too, even with its best possible performance and said post poll alliance, we expect UPA to fall short of 272 majority.

The vote share irony: Our calculations indicate a ~36% vote share for NDA (ex-TDP) which translated to 318 seats (ex-TDP). Interestingly, C-voter is expecting NDA to garner 287 seats in 2019 election (lower than that of 2014) with an increased ~42% vote share though. This difference could be emerging from asymmetric distribution of electorate and conservative estimates by C-voter, but this an interesting statistic nonetheless.