Election Surprise Edition

Dear Senderoistas,

As I am busily involved with our annual migration south, I hadn't anticipated sending another Sendero so soon.

I am only doing so due to my coming across a very comprehensive piece of research on the coming US presidential election, forwarded to me by my friend and colleague Stephen McBride the firecracker behind our RiskHedge.com service.

The research was done by Mark Bower on behalf of Jawad Mian's Stray Reflection service. Jawad, who is based in the United Arab Emirates, shares my affection for Arabian horses and together we have mused about a ride-around somewhere in the Middle East, something I would enjoy very much.

In any event, unlike the main stream news, which increasingly appears willing to shred it's own credibility in the quest to take down President Trump, Mark does the hard work of picking apart the various polls and historically important indicators of election success.

In doing so, he now believes that President Trump will prevail in the imminent election.

Personally, I have been placing some wagers on the outcome, and now up to $3,100 in bets.

It is funny to observe the fellows I am betting against, literally high-fiving each other in glee at the opportunity of collecting easy money.

If they had the benefit of Bower's insights they wouldn't be so quick to count their chickens.

Mark's research follows, in two parts, after which I'll have a quick parting comment. As you'll see, Mark's research is extremely comprehensive, the most comprehensive I have seen.

As background, Mark is a professional investor and an advocate of Collective Equity Ownership.

He has an extensive background of almost 20 years on Wall Street. He was the Head of Emerging Markets for Bienville, a $2.5bn multi family office in New York.

Before that, he researched equities for Argonaut and Prince Street Capital Management. Mark also serves as a private investor in over a dozen companies in a variety of industries and geographies. He was an early investor in Qulia Labs, Opportunity Network, Medefair and 84 Collective in Vietnam. Finally he serves as the Chief Commercial Officer for Life & Soul Pictures in London.

In case you find yourself in the need of some calming, yet lively, music to calm the soul in these trying times, I recently recalled an album I used to listen to while writing. It's called Deep Forest, and you can listen to here.

2020 Election Thoughts - Part I

Mark Bower

October 2020

I am going to lay out evidence for five points of my election thesis:

1) The market and mainstream press expect ‘a bad outcome’: a Biden win, Dem control of Senate, and/or a contested election that goes to the courts, like in 2000

2) Contrary to popular opinion, Trump is highly likely to win the popular vote, unlike in 2016, because of several factors:

a. Key pollsters are showing a stronger trend already

b. Trump has much higher support among minority voters than in 2016

c. Democrat and Independent enthusiasm for Biden/Harris is low

d. Several things that correlate to popular vote suggest strong standing

3) Trump will almost certainly win the electoral college, as nearly all the key states have factors that will keep them voting for Trump, and likely expand his majority in states that were close in 2016

4) The Dem reliance on mail in voting is a big factor- it will hurt their overall vote totals, and it will show a Trump **tsunami* on election night that will shock the market when trading reopens as every short has to cover

5) The outcome in markets should be similar to 2004 when George Bush defeated John Kerry and you saw a giant rally or 2016 when Trump won the surprise election

1) The market and mainstream press expect a bad outcome, and markets are showing higher implied vol for post election:

The polls show a +7 lead by Biden (this is an aggregate of polls from realclear politics)

The famous Nate Silver, of 538, showing a 77% chance that Biden wins.

Betting markets show Biden winning too, although by a narrower margin.

The markets are worried about any negative outcome. For example, S&P implied vol for November/December has a 3 or 4point ‘pop’ (so not just for election night, but also hedging for an outcome similar to 2000- a contested election tied up for two months in the courts).

VIX curve is showing elevated premia (c/o Bloomberg: https://www.bloombergquint.com/markets/u-s-markets-gird-for-volatility-long-after-election-day-passes)

Headlines abound with awful outcomes of courts, military and contested election:

Could Election be militarized: https://www.seattletimes.com/nation-world/nation-politics/at-pentagon-fears-grow-that-trump-will-pull-military-into-election-unrest/

A Nightmarish 2020: https://nymag.com/intelligencer/2020/09/a-nightmarish-end-to-2020-is-becoming-all-too-plausible.html

What if Trump refuses to concede: https://www.theatlantic.com/magazine/archive/2020/11/what-if-trump-refuses-concede/616424/

What does a contested election mean for American investors? https://www.cnn.com/2020/09/24/investing/premarket-stocks-trading/index.html

2) In 2016, Trump lost the popular vote to Hilary Clinton, but, won in the electoral college [more on that later]. This time, I think it is likely that Trump will win the popular vote, and that will almost guarantee an electoral college win.

2016 Popular Vote

Candidate

Percent of Vote

Clinton

48.2%

Trump

46.1%

Johnson

3.3%

Stein

1.1%

McMullin

0.5%

One polling company, Rasmussen had the result exactly correct in 2016 while some were wrong by almost 10%.   What do they show now? Basically a tied outcome in popular vote, which will lead to an electoral college win.

“The new national telephone and online survey finds Biden leading President Trump 48% to 47% among Likely U.S. Voters. Three percent (3%) prefer some other candidate. Just two percent (2%) are undecided. (To see survey question wording, click here.)”

Rasmussen also runs a poll that is a daily presidential approval index. They index this to Obama’s first term of 2009-2012. Trump has been above 50% in September. If this plays out in a similar fashion to Obama’s first term, Trump should be re-elected with a small popular vote majority and 320-340 electoral votes (270 to win).

A second way to look at the popular vote likely outcome that worked in 2016 by state and nationally is to find out how many Democrat, Republican and Independent voters there, find how likely they are to vote for each candidate, and do some simple math. The 2016 polls were quite accurate in that they knew what % of each party voted for each candidate, but, their samples were wildly off. They nearly always oversampled Democrats, and the numbers would routinely reach 8-10% too many Dem voters in their sample.

One pollster that frequently does this is Gallup. The recent data from Gallup shows equal amounts of Democrats and Republicans, and around 40% identifying as Independents:

Republicans

Independents

Democrats

%

%

%

2020 Aug 31-Sep 13

29

40

30

Given the Rasmussen numbers on which way Republican, Independent, and Dem voters identify, that should be a 1-2% popular vote victory for Trump:

The new survey finds Trump with 83% support among Republicans and a six-point lead among voters not affiliated with either major party. Biden has 83% of the Democrat vote.

If you weight the popular vote total after those numbers by numerator and denominator you get the following popular vote data:

 

% of Electorate

% for Trump

% for Biden

Total % for R

Total % for Ds

Republicans

29

83

17

24.07

4.93

Independents

40

53

47

21.2

18.8

Democrats

30

17

83

5.1

24.9

Total Votes

50.37

48.63

 

Now let’s look at a recent poll from A rated Pollster Monmouth: It had 50% Biden, 44% for Trump. But the breakdown of voters was 29% R , 36% Independent 35% Democratic. Reweighting the same poll by the Gallup numbers effectively gets a tie in the popular vote. So, again, the pollsters are somewhat accurate but their samples are wrong.

 

% of Electorate

% for Trump

% for Biden

Total % for Trump

Total % for Biden

Republicans

29

96

4

27.84

1.16

Independents

40

43

41

17.2

16.4

Democrats

30

2

95

0.6

28.5

45.64

46.06

 

Demographic breakdown

 

One key factor is that Trump’s support among minority/ non white voters, who are 30% of the electorate, are up significantly versus 2016. In 2016, he had very poor support among African Americans, and not good support among Asians and Hispanics.

 

Clinton

Trump

Other Candidate

% of Total

White

37

58

5

70

Black

88

8

4

12

Asian

65

29

6

4

Other

56

37

7

3

Hispanic

65

29

6

11

Polling in 2020 shows much higher support.

Trump 2016

Rasmussen Data

White

58

54

Black

8

32

Asian

29

47

Other

37

Hispanic

29

So a 10% shift in 30% of the electorate means a 3% popular vote advantage for Trump, and a victory in the popular vote, assuming similar white numbers and similar turnout figures.

Running those numbers, assuming similar turnout and similar numbers for non minorities, gives you a 2% win in the popular vote and 312 electoral votes according to Cook

Biden Votes

Trump Votes

Trump Margin

US Total

65,381,572

68,440,731

2.2%

3) Trump won the electoral college in 2016, and will almost certainly win again in 2020.

First of all, what exactly is this ‘electoral college’ and why does it matter?

The US, unlike nearly every other country with a presidential system, does not allocate the job to the person who gets the most votes from the population. Instead, it has what it is essentially a points system – each state gets points equivalent to its number of Representatives in Congress and number of Senators.

On top of that, Washington DC, which has no senators or Congress representatives, gets 3. With a total of 100 Senators, 435 Representatives, and 3 for DC that means there are 538 ‘points’ total. If a candidate wins the most votes in a state, he gets assigned all of the points from that state [with a couple of tiny exceptions]. If a presidential candidate gets 270 of these points, he or she has a majority and wins the election.  In  a two party system, it’s a race to see who can get enough states to 270 votes first – there is no coalition system possible like in other countries.

Why does the US do things this way? It’s complicated, but, essentially, the theory during the founding of the US was that it was better to weight the outcome of the presidential election more to the smaller states, which have may have 5% of the population of the bigger states but the same number of Senators. That way, the five most populous states have less ability to use their mass to win a small majority of votes nationally and impose their agenda on the other 45 states. In practice, this has actually been the case- in 2000 and 2016, the popular vote winner did not win the electoral college totals.

In 2016, Trump lost the popular vote, as voters from large states like California, Illinois and New York voted overwhelmingly against him, but he gained a plurality in enough states to win 306 points- much more than the 270 needed to win.

Many states are reliably Democratic [California, Oregon, New York], and many are reliably Republican [Kansas, Idaho, Tennessee, Texas] and so the election really revolves around the points from 10-15 states that have voted for both parties.

In 2016, Trump was able to turn several such states, like Florida [29 points] and Ohio [18] and he was able to finally win three states that Democrats had won each time since 1988 – Wisconsin [10], Michigan [16], and Pennsylvania [20]. Those final three he won by less than 2% each, but, those points caused him to win the election.

Right now, this is a list of the states that are considered ‘likely Biden’ [with 218 points], with likely Trump states [169 points] and the ten states in the middle [151] points that will determine the election.

Here is what happened in 2016- Trump won 8/10 states, and a total of 135 electoral points.

 

 

Electoral Points

Trump 2016 %

Clinton 2016 %

Difference

Arizona

AZ

11

48.7%

45.1%

Trump +3.6

Florida

FL

29

49.0%

47.8%

Trump +1.2

Georgia

GA

16

50.8%

45.6%

Trump +5.2

Michigan

MI

16

47.5%

47.3%

Trump +0.2

Minnesota

MN

10

44.9%

46.4%

Clinton +1.5

New Hampshire

NH

4

46.6%

47.0%

Clinton +0.4

North Carolina

NC

15

49.8%

46.2%

Trump +3.6

Ohio

OH

18

51.7%

43.6%

Trump +8.1

Pennsylvania

PA

20

48.2%

47.5%

Trump +0.7

Wisconsin

WI

10

47.2%

46.5%

Trump +0.7

What is the outlook for those states in 2020? Well, the polls seem to indicate that Trump should win 9 out of 10 or 9 out of 10, giving him 316 to 320 electoral votes, easily a majority. The most accurate pollster for state races in 2016 and 2018 was Trafalgar Group. I put their numbers together:

 

 

Electoral Points

Trump 2020 %

Biden 2020 %

Difference

Arizona

AZ

11

46.2%

44.8%

Trump +1.4

Florida

FL

29

48.7%

45.6%

Trump +2.9

Georgia

GA

16

49.8%

43.3%

Trump +5.5

Michigan

MI

16

46.7%

46.0%

Trump +0.7

Minnesota

MN

10

46.5%

46.9%

Biden +0.4

New Hampshire

NH

4

-

--

-

North Carolina

NC

15

47.8%

46.1%

Trump +1.7

Ohio

OH

18

-

-

-

Pennsylvania

PA

20

45.9 %

48.1%

Biden + 2.2

Wisconsin

WI

10

46.2%

45.4%

Trump +0.8

In Ohio and New Hampshire there have not been reliable polls by Trafalgar, but Trump won Ohio by 8% last time and so I will assume that goes his way again. Trump is up comfortably in several swing states [Ohio, Georgia], competitive in all of them, and has been adding support versus the last election in Minnesota, Wisconsin and Michigan. All of these states have typically voted Democrat, so Biden is in deep trouble in his heartland.

This list shows that Biden would only add potentially 34 electoral votes, and that’s if he hangs on in Minnesota and wins New Hampshire. That only gets him to 252, 18 votes short. So he will need to flip a big state back from Trump – Florida - or, win several small states to win the presidency. However this list indicates that is extremely unlikely, as he is behind in 7 out of 10. If Biden cannot win Pennsylvania, it just won’t happen for him.

Making this even more remote for Biden is the fact that voter registration in the past four years has heavily favored Republicans.   Several states provide party registration data of votes and three swing states stand out:

· In North Carolina, Republicans have added 120k voters since 2016 and Democrats have lost 119k voters

· In Florida, Republicans have added 150k more voters than Democrats over the past four years

· In Pennsylvania, Republicans have added 125k voters while Democrats have lost 35,000 voters

ACG analytics did a comparison of how the electorate in several states has shifted due to voter registration. The percentage of Democrat voters has declined in several states that Biden has to win, most notably North Carolina and Pennsylvania.

Finally, running a demographic model with slightly higher minority support for Trump gets you 312 electoral votes – I just moved African Americans 15% more to Trump, Asians and Hispanics 10%, which is still below what national polls are showing.

And while the polls may say one thing among likely voters, it is very likely that Democratic turnout will disappoint. Frankly, the Democrats’ presence both in person and online is very poor relative to the Republicans’. The Biden campaign decided not to knock on doors and meet people to motivate voters, owing to Covid-19 fears:

From CNN:

Trump Victory, the joint operation between the RNC and the Trump campaign, has an army of 2,000 paid field staffers in 17 states and more than 2 million volunteers making phone calls and knocking on doors. The field operation claims to have made more than 90 million voter contacts in the cycle, including 12 million door knocks since they resumed the practice in June.

In just the last week, according to Trump Victory spokesman Rick Gorka, volunteers have knocked on more than 533,000 doors across the key states of North Carolina, Michigan, Wisconsin, Minnesota and Georgia.

From Politico:

Donald Trump’s campaign says it knocked on over 1 million doors in the past week alone.

Joe Biden’s campaign says it knocked on zero.

From Vox:

Politico’s Alex Thompson reported in early August that the Biden campaign was knocking on zero doors each week, while the Trump campaign claims to have knocked on over a million. More recently, in Michigan, the Trump campaign boasts that it has “an army of over 43,000 volunteers and staff covering all 83 counties,” whereas the state Democratic Party told reporter Jonathan Oosting it was “not yet comfortable” knocking on doors. On August 28, the Trump campaign bragged about knocking on its 1 millionth door in Florida. Door-knocking numbers like this aren’t really useful (they don’t tell you how many people were actually contacted), but the pattern is illustrative.

The Biden campaign is responding to the risk that door-to-door canvassing will spread Covid-19 infection by shutting down its door-knocking efforts.

So, you may say, the Trump campaign has the on the ground presence, but surely Biden will win the game online given their focus there? Well, a comparison of social media accounts shows that on every single platform, Trump has 5 to 10x the followers:

# of followers

Trump

Biden

Twitter

86.2m

9.6m

Facebook

31.1m

2.9m

Instagram

22m

4.1m

Youtube

1.1m

194K

The only possible exception is that pro Biden CNN has 10.7m YouTube followers , while Fox News has 6.3m followers.

4) Mail in Voting and its Implications

The Democrats had heavily pushed mail in voting, to make sure that everyone could participate without fear of getting COVID. Polls consistently show 15-20% more Dems will vote by mail. One CNN poll had half of Dems voting by mail.

However, this is going to create a huge problem – nationally, 2 to 4% ballots get rejected ! So the likeliest is that in addition to lower Dem turnout from the above factors, that Dem voters will miss the deadline to send a ballot in. Some examples from Florida and Pennsylvania:

In Pennsylvania, for example, election officials have warned that they will reject any so-called “naked ballots” that do not arrive in a “secrecy envelope” that ensures anonymity for each vote. Some have said that up to 100,000 votes could be tossed out. That is an alarming possibility, particularly given that Pennsylvania is likely to prove pivotal in this year’s election. (WAPO)

Why does that matter? Well in Pennsylvania this year, 2.25m people have requested mail in ballots, 64% of them Democrats and 26% Republicans. This is in a state Trump won by 70k votes last time. If 100,000 votes aren’t counted, then the wide majority of them will be Democrat votes, and that will further help Trump.

Florida threw out 1.5% of ballots in the primaries. That’s a huge number across the whole state:

https://www.politico.com/states/florida/story/2020/09/17/more-than-35-000-mail-in-ballots-were-rejected-in-florida-primary-1317327

“This could be a huge problem in November,” Smith said. “We could exceed 100,000 vote-by-mail ballots that don’t count.“

But the majority of those getting tossed, are going to be Democratic ballots.

Over 3.2 million active registered voters in Trump’s home state of Florida had requested absentee ballots as of Friday evening….About 46% of those active voters requesting ballots so far are Democrats, while about 31% are Republican

The Dems have been trying to change back, but, this is a very risky move with only one month to go.

Finally, these votes can often arrive on or after Election Day, and take up to 12 days to get counted. So if 20 to 30% more Democrats are voting by mail, then the Trump lead in the popular vote instead of being 2% on election night, is going to initially look like 6-9% ! That’s as the in person votes (for Republicans) get counted that day and the mail in votes (for Democrats) get added up over the following week. And many of those votes will be rejected for some technical reason.

Instead of guaranteeing high turnout and an election win, the mail in voting strategy is going to cause a very poor result for the Dems relative to their polling numbers.

5) Market Reaction

Nomura had piece on the technical potential of an up-crash that could result from the combination of a better than expected outcome, half from technical factors of mail in voting and half from support for Trump larger than what the polls show.

https://www.zerohedge.com/markets/crash-risk-what-if-election-not-doomsday-event-vix-expects-it-be

After the 2004 election, the US broad indices rallied strongly, as did emerging markets- it was the reflation trade (Nov 2ndelection night)

SPY

Date

Open

High

Low

Close*

% Change

4-Nov-04

114.78

116.67

114.68

116.55

2.6%

3-Nov-04

115.03

115.36

114.24

114.98

2-Nov-04

113.67

114.57

113.22

113.55

1-Nov-04

113.56

113.84

113.2

113.51

EEM

Date

Open

High

Low

Close*

% Change

4-Nov-04

20.21

20.57

20.18

20.5

3.1%

3-Nov-04

20.36

20.39

20.17

20.19

2-Nov-04

19.94

20.1

19.89

19.9

1-Nov-04

19.88

19.89

19.76

19.86

EWZ

Date

Open

High

Low

Close*

% Change

4-Nov-04

19.3

19.48

19.27

19.43

4.1%

3-Nov-04

19.35

19.37

19.17

19.2

2-Nov-04

18.99

19.2

18.93

18.95

1-Nov-04

18.67

18.94

18.66

18.93

XLE

Date

Open

High

Low

Close*

% Change

4-Nov-04

35.19

35.6

35.19

35.52

0.5%

3-Nov-04

35.1

35.25

34.6

35.1

2-Nov-04

34.8

34.93

34.32

34.41

1-Nov-04

35.35

35.52

34.71

34.93

IWM

Date

Open

High

Low

Close*

% Change

4-Nov-04

59.28

60.03

58.92

59.99

3.0%

3-Nov-04

59.21

59.55

58.86

59.26

2-Nov-04

58.63

59.03

58.07

58.33

1-Nov-04

58.27

58.63

57.86

58.56

After 2016 (November 8th election date), the broad US indices rallied strongly, led by IWM (small caps). Emerging markets sold off on the fear that the new Trump administration would be protectionist, harsh on China, and lead to a stronger dollar.

SPY

Date

Open

High

Low

Close*

% Change

10-Nov-16

217.3

218.31

215.22

216.92

4.0%

9-Nov-16

212.37

217.1

212.34

216.38

 

8-Nov-16

212.69

214.77

212.38

214.11

 

7-Nov-16

208.55

213.19

208.55

213.15

 

EEM

Date

Open

High

Low

Close*

% Change

10-Nov-16

35.92

36.05

35.04

35.22

-4.3%

9-Nov-16

36.34

36.69

36.09

36.25

 

8-Nov-16

37.23

37.67

36.97

37.47

 

7-Nov-16

36.81

37.27

36.75

37.23

 

EWZ

Date

Open

High

Low

Close*

% Change

10-Nov-16

34.51

34.86

33.14

33.45

-9.0%

9-Nov-16

37.54

37.54

36.1

36.31

8-Nov-16

36.88

37.88

36.65

37.54

7-Nov-16

36.75

37.27

36.67

37.27

XLE

Date

Open

High

Low

Close*

% Change

10-Nov-16

70.4

71.28

70.16

70.75

3.2%

9-Nov-16

69.56

70.97

68.75

70.45

8-Nov-16

68.89

69.71

68.79

69.32

7-Nov-16

68.54

69.27

68.54

69.23

IWM

Date

Open

High

Low

Close*

% Change

10-Nov-16

124.26

125.4

122.62

124.5

5.5%

9-Nov-16

118.08

122.83

118.05

122.54

8-Nov-16

118.3

119.46

117.71

118.88

7-Nov-16

118

118.82

117.71

118.57

This time around, with those two elections as a backdrop, I think three things will happen:

1. Trade: Buy XLE and XOP (Energy and Energy Services ETF’s) Energy is the worst performing sector on the year, partially on the Q1 crash in oil prices, and partially on the worries of a Biden administration. The Biden plan is to ban fracking and heavily tax energy companies. In the event of a Trump win, I think we would see powerful rotation out of tech into energy stocks.

2. Trade: Buy IWM (US Small Caps). The last two times that a Republican candidate had a surprise win or a win in a close election, small caps rallied powerfully – 3.0% and 5.5%. Given their underperformance, we should see a rotation.

3. Trade: Sell December Vol- SPY Implied vol is elevated until the end of the year. As certainty is restored, not only should near term vol, but, the further out vol where people are hedging themselves against a contested election should fall the most.

As for emerging and international markets? I think the prospect of a 2016 crash in EM is off the table, as Trump has largely implemented his realignment of trade policies, and apart from China, has no new aggressive measures. The two factors will be investors assuming higher growth versus investors’ expectations of a stronger dollar. My guess is that the growth narrative would win out slightly and investors under-positioning would cause them to play catch-up, but no strong conviction here.

2020 Election Thoughts - Part Two

Mark Bower

October 20th 2020

A few weeks has passed since the initial election thoughts were put together. Where are we now? After a Supreme Court battle, Trump getting COVID, an ugly first debate, and all sorts of news, I think that overall the data is strengthening for Trump, and it’s being masked by anomalies in the national polls which I will address.

"But how can Trump win when major polls show him down 8, 10, 14 points?” will be the key question to answer, and the answer has three parts. The first point is that the vast majority of pollsters do not correctly sample the electorate, and adjusting their results for the real electorate shows a close race nationally.

Secondly, results revolve heavily around turnout- and both party registration and early voting data show that Republican turnout is high, and Democrat turnout is weak. Thirdly, since this is determined by the electoral college, and the ten ‘battleground states’ we have to look at the specific results as opposed to just a national poll. In those battleground states, Trump is ahead in seven out of ten, and Biden’s largest state lead is 1.3%. If turnout continues to run low, Biden could easily lose all ten battlegrounds and hence the election.

1) What are current polls and current market saying - showing a huge Biden victory

2) A breakdown of what I am seeing beneath the surface of major national polls and their inconsistencies:

a. Approval data remains favorable to Trump

b. Party Registration Data shows a lead by Republicans

c. Internals of national polls

d. Properly weighing polls gets surprising outcomes

3) State by state update - comparison to 2016 path, latest Trafalgar data

4) Early voting update

5) Calendar

6) Predictions!

a.      Trump wins popular vote by 1.5-2.5%

b.     Trump wins electoral college easily, with 316-332 points

c.      Republicans keep the Senate

d.     Some ‘safe Dem’ states will be surprisingly competitive

1) National poll evolution last month: All the major outlets are showing a big Biden lead and the market is now pricing in a big Democratic victory.
Real Clear Politics average has Biden up 9%:

Nate Silver at 538 has an 87% chance that Biden wins.

Betting markets are at 65% chance, after viewing this as a tie two months ago:

2) National Polls and breaking them down. The national polls are showing an amusing outcome- Biden up by 10%. As a reminder, this is an even larger margin than what Obama won with in his historic 2008 victory. The 538 national average shows that this has remained steady the past six months:

On the other hand, let’s look at some things that predict popular vote. First up is Trump’s approval rating, which Rasmussen tracks daily. It has held in the 48-50% range for the past month, exactly where Obama was in 2012 before his re-election victory by 3 or 4%.

So why do I think that the outcome will not be Biden + 10, but instead Trump +2 (and an electoral college victory) ? Because the sample is wrong. The voters participating in these polls do not match the electorate. In order to determine the likely outcome, we need to in essence take their numbers on party ID and who is voting for which candidate, but make it resemble the electorate.

The latest Gallup party ID data shows a 1 percentage point lead by Republicans in party registration over Democrats and a lot of independents. Gallup was within 1 or 2% of the actual result in the last four cycles, so this is a very good guide.

Look closely at the trend since 2016- R’s holding steady at 28% while D’s are down from 30% to 27%. Clinton won the popular vote by 2%, so a 3% shift in the registration to Republican gets you a roughly 1% victory in popular vote for Trump depending on how independent voters vote.

Moving on - the two most accurate national pollsters in 2016 were Rasmussen and IBD/TIPP. (The third was USC Latimes, but they have changed their methodology).  Rasmussen’s last poll was from a week ago and I could not get the more detailed breakdowns for it. As a substitute here is the latest IBD data, as Rasmussen and IBD closely track each other. This data, as of October 20th, is showing Biden ahead by 2%. Here is party breakdown of how R’s D’s and Ind’s will vote.

Trump

Biden

Party

 

 

Democrats

7.40%

89.90%

Republicans

94.30%

3.80%

Ind./Other

40.60%

44.70%

Now if we take that data, and match it to the Gallup data on party ID – what percentage will each candidate receive? It turns out that you get a result of a 1.5% percent popular vote victory for Trump.

% of Electorate

% for Trump

% for Biden

Total % for Trump

Total % for Biden

 

Republicans

28

94

4

26.32

1.12

 

Independents

42

41

44

17.22

18.48

 

Democrats

27

7

90

1.89

24.3

 

 

45.4%

43.9%

 

Now let’s at the NYT Siena poll released today, their final one. It shows Biden ahead by 9 points (50-41). They are not exactly known for being favorable to Trump or Republicans, and their last poll of 2016 was a huge miss, with Clinton winning by 46 to 39 over Trump.

Looking at the internals of the polls, we can see their sampling is way off of the population and the electorate with Democrats outnumbering Republicans by six percentage points:

If we take their data and adust it to Gallup, what do we get? A close race, with popular vote numbers exactly like 2016.

% of Electorate

% for Trump

% for Biden

Total % for Trump

Total % for Biden

Republicans

28

91

4

25.48

1.12

Independents

42

37

46

15.54

19.32

Democrats

27

7

93

1.89

25.11

42.9

45.5

I would also add that their numbers are probably too low for Trump because of several anomalies. The two largest are the Trump votes by African Americans was at 4% in this poll, and Hispanics at 30%. Those are lower than any Republican candidate has ever gotten, and, stand in direct contradiction to approval data, early voting and state by state level data.

If we change the African American number from 4% to 15%, and the Hispanic number from 30 to 35% then we’d get a tie in the popular vote. And this is coming from a pollster that was off in 2016 by five percentage points.

Hence why I am so confident in strong Trump showing in the popular vote: the polling sample does not match the electorate. When you look at the actual people voting on November 3rd, the data looks much different.

3) Early Voting Data. All states in the US allow you to vote before election day, either in person or by mail. Given COVID, numbers of people voting early are up over 2016,  by almost 300%. In addition, Democrats banked on this strategy to drive turnout. How does it look now?

In essence, not great. Democrats requested 47% of ballots versus 30% for Republicans, but the returned ballots are now running at 48% vs. 34%. In other words, Republicans are showing up in higher percentages than Democrats.

(source NBC/Target Smart)

Another key variable is that the group which is mailing in the fewest ballots is the 18-29 group, who typically vote 60-70% Democratic. Here is the early vote data by age group, showing a big shortfall for younger voters:

(source NBC/Target Smart)

4) State by State Outlook. Many have looked at the state by state data and the Real Clear Politics Battleground average and are saying Trump has no chance. Here is a graph of the Swing State Average from RCP:

Even the Biden campaign thinks these numbers are wrong[1]

Also, we need to look back to see how off the 2016 polling data was. One Twitter analyst put together four states to show the October 2016 and October 2020 polling averages….there was a huge miss by major polls. Only Trafalgar and one or two others were able to successfully call the swing states, Michigan most prominently.

With that as a background, we return to the Electoral College and the points from each state. Remember, that most states can safely be classified for one of the two parties. This time around, those safe states generate approximately 218 electoral votes for Biden and 169 for Trump [270 to win]. Neither candidate has enough points to win with just their ‘safe states’, so the election will be determined by the outcome in those states.

In 2016, those ten states were very competitive, and many were determined by razor thin margins.

 

 

Electoral Points

Trump 2016 %

Clinton 2016 %

Difference

Arizona

AZ

11

48.7%

45.1%

Trump +3.6

Florida

FL

29

49.0%

47.8%

Trump +1.2

Georgia

GA

16

50.8%

45.6%

Trump +5.2

Michigan

MI

16

47.5%

47.3%

Trump +0.2

Minnesota

MN

10

44.9%

46.4%

Clinton +1.5

New Hampshire

NH

4

46.6%

47.0%

Clinton +0.4

North Carolina

NC

15

49.8%

46.2%

Trump +3.6

Ohio

OH

18

51.7%

43.6%

Trump +8.1

Pennsylvania

PA

20

48.2%

47.5%

Trump +0.7

Wisconsin

WI

10

47.2%

46.5%

Trump +0.7

How are those states looking in the current race? According to one of the most accurate 2016 and 2018 pollsters, Trafalgar, they are again looking close, but Trump has a distinct edge.

 

 

Electoral Points

Trump %

Biden %

Difference

Arizona

AZ

11

47.8%

43.8%

Trump +4.0

Florida

FL

29

48.4%

46.1%

Trump +2.3

Georgia

GA

16

49.8%

43.3%

Trump +5.5

Michigan

MI

16

46.5%

45.9%

Trump +0.6

Minnesota

MN

10

46.5%

46.9%

Biden +0.4

New Hampshire

NH

4

-

-

-

North Carolina

NC

15

47.8%

46.1%

Trump +1.7

Ohio

OH

18

47.6%

43.9%

Trump +3.7

Pennsylvania

PA

20

46.4%

47.5%

Biden + 1.1

Wisconsin

WI

10

46.3%

47.6%

Biden + 1.3

There is no polling data yet on New Hampshire from Trafalgar.  The outcome of these races show total additional Electoral Points of 115 for Trump and 44 for Biden (assuming he wins New Hampshire where Dems have a party advantage). That makes the totals at 284 for Trump and 262 for Biden, in other words, a Trump victory.

To emphasize what the math is telling us, it appears that Biden is extremely vulnerable. Whereas Trump leads several states by a comfortable margin, such as Florida [29], Ohio [18] and Georgia [16],  the three states with a Biden lead are at 1% or less. A failure to turn out his base that’s been suggested in the early voting data would mean that he could lose nine or ten out of the battlegrounds, a decisive loss in the electoral college. If Trump wins all these battleground states, his total would come out to 318 votes, larger than his 2016 victory.

5) Calendar

The upcoming calendar has one debate and a lot of campaign stops before November 3rd Election Day. I imagine that the Trump team will have some cards up its sleeve ahead of the final debate, trying to publicly pin Biden to the wall.

Also note that the Electoral College representatives will have to be picked by December 14th- meaning, states will have to determine who won the votes from that state.

So big court challenges will have to be sorted out in the five weeks after election day.

Date

Event

October 22nd

Final Presidential Debate

November 3rd

General Election Day

December 8th

Preliminary vote by Electoral College representatives

December 14th

Final vote by Electoral College representatives

January 20th

Inauguration Day

6) Predictions!

a. Trump wins Popular Vote by 2.0%

b. Trump wins electoral college- base case is 316-322 range, bear case 296, bull case is 331 [see below]

c. Republicans hold the Senate – 52 or 53 seats


I have put together my predictions on swing states. I think that Trump will carry every state he carried in 2016, with the addition of Minnesota and Nevada.

The strong voter registration gains will expand his base in Ohio, Pennsylvania, Florida, North Carolina.

Low Dem turnout should allow him to again squeeze out Wisconsin and also to win Minnesota.

Higher minority support will sure up Florida, Ohio and Pennsylvania too.

Arizona and Georgia were never really up for debate in this cycle – only in their Senate races.

New Hampshire, Clinton won by a narrow margin, and has seen 3.5% shift towards Dems in registration, so I think Biden will win. But it’s a state with only 1.4m people, so it’s much harder to get accurate data and decent polls.

 

 

Electoral Points

Trump 2016 %

Clinton 2016 %

Prediction

Arizona

AZ

11

48.7%

45.1%

Trump +3.0

Florida

FL

29

49.0%

47.8%

Trump +4.0

Georgia

GA

16

50.8%

45.6%

Trump +5.0

Michigan

MI

16

47.5%

47.3%

Trump +1.5

Minnesota

MN

10

44.9%

46.4%

Trump +1.0

New Hampshire

NH

4

46.6%

47.0%

Biden +2.0

North Carolina

NC

15

49.8%

46.2%

Trump +5.0

Ohio

OH

18

51.7%

43.6%

Trump +9.0

Pennsylvania

PA

20

48.2%

47.5%

Trump +2.0

Wisconsin

WI

10

47.2%

46.5%

Trump +1.5

Four electoral surprises – these are all ranked as safe Dem states, but, I think some surprises will come. It’s very possible that Trump flips either New Mexico and Nevada - I am guessing it will be Nevada given Trump’s higher Hispanic support, and party registration losses of 1.5% by the Dems since 2016.

In Colorado, I think higher Hispanic support and no third party candidate mean that Trump will outperform, although it still goes to Biden.

With regards to Oregon, the early data shows essentially two states- Portland area is heavily Democratic and pro-Biden, but early voting in many counties outside of Portland is showing very high Republican turnout.  Biden should win, but it won’t be by a huge majority.

 

 

Electoral Points

Trump 2016 %

Clinton 2016 %

2020 Prediction

Colorado

CO

9

43.3%

48.2%

Biden +2.0

Nevada

NV

6

45.5%

47.9%

Trump +1.0

New Mexico

NM

5

40.0%

48.3%

Biden +2.0

Oregon

OR

7

39.9%

50.1%

Biden +4.0


[1] https://nypost.com/2020/10/18/biden-campaign-memo-warns-trump-neck-and-neck-in-several-key-states/?utm_source=twitter_sitebuttons&utm_medium=site%20buttons&utm_campaign=site%20buttons

Parting Thoughts

David here again.

An anecdote from on the ground here in the liberal bastion of Vermont.

Our house here is a tick over 6 miles from the center of the nearest small town. From that town to a golf course in the next town over where I play fairly regularly is about 20 miles.

Yesterday, my business partner and dear friend Olivier and I drove about an hour in the opposite direction to another town, Randolph, to play a final game of the season.

Despite the election being imminent, in all those miles I have only counted a total of six "Biden for President" yard signs. In other words, the lack of enthusiasm for Biden as a candidate is, my experience, unprecedented.

My dear mother, a life-long Democrat (despite her protestations), told me the other day that she had voted early for "the tin can". To wit, anyone other than Trump.

So what do you think the less dedicated voters will do? For example, the much touted liberal-leaning youth vote, especially now that the universities, the usual gathering spot for political organizers to whip up crowds, are all but deserted? Or, what about the minority populations, will they really look at the stumbling white socialist sock puppet with affection and enthusiasm?

Will they really turn out in droves for the "tin can"?

I won't be spending my potential $3,100 winnings just yet, but despite the media blowing their party horns loudly to herald in the political demise of Trump, this race is far from over.

And with a very special thanks to Mark Bower, I will sign off for now, making my final preparations before hitting the trail back to Argentina.

Until next time,

David

Subscribe to our mailing list

* indicates required