Thursday, March 12, 2020

Wilshire 5000 Under Trump

The Wilshire 5000 is a broad measure of the net worth of the entire US stock market -- it's numerical value is the market's value in billions of dollars:


It's now down $9.7 trillion from its maximum.

Here's its record under recent presidential administrations:


Clearly in the last 27 years the market has done far better under Democratic presidents than under Republican presidents. It's not even close.

(Before this latest market crisis the W5000 was up about the same under Trump 1 as it was under Obama 2 (in its average percent increase per year (last column)).) 

17 comments:

Layzej said...

My sense is that Trump's tax cuts funded some portion of the rise, and that there would have been a fall regardless of how he handled the COVID case.

There will need to be a stimulus package to pull the market out of this dip. My big fear is that Trump already spent his stimulus money. The US deficit is already at about the levels seen during the stimulus spending following the 2008 crisis. Maybe there is still plenty of room to grow? I don't know.

David in Cal said...

Clearly in the last 27 years the market has done far better under Democratic presidents than under Republican presidents. It's not even close.

It's silly to attribute an significance to this statement, for several reasons.

1. The sample size is ridiculously small

2. The results are enormously affected by specific, unique events, for which the President doesn't deserve blame or credit, such as the high tech boom and the current pandemic.

3. Crediting Obama with results through 1/20/2017 ignores the big rise in the stock market between November, 2016 and Jan, 2017, which was due to Trump's election. The Dow Jones rose about 11% during this period. That amount should be deducted from Obama's record and added to Trump's.

Cheers

David Appell said...

David, if Trump gets credit for confidence instilled immediately after the Nov 2016 election, why doesn't Obama get credit for the long recovery after the 2008 financial crisis?

Republicans also run higher deficits than Democrats. It's Democrats who bring them down to a reasonable size after Republicans run them up.

Layzej said...

Point 3 is quite a stretch, especially following on point 2.

David Appell said...

"There will need to be a stimulus package to pull the market out of this dip. My big fear is that Trump already spent his stimulus money. The US deficit is already at about the levels seen during the stimulus spending following the 2008 crisis."

Great point that I've been thinking about too. But I don't think Trump would hesitate to run a deficit of $2 trillion/yr if it helped his reelection chances. (Current deficit is about $1 T in the last 12 months.)

The boom years are when you're supposed to be paying the debt down, so you have credit when you need it. At least that's what Keynes said. Not a lot of Keynesians in the GOP.

David Appell said...

But I'll admit it's a little unfair to post this graph just after the market has taken a big hit -- cherrypicking of a sort. But Trump has talked about the market as if he gets all the credit for it, so then he gets the blame when it crashes.

Ned said...

Starting with Clinton might be considered cherry-picking. I downloaded the data from FRED, which has them starting in 1970. Here are the fractional changes during each 4-year presidential term (1.00 = no change), from worst to best:

GW Bush_2 ...... 0.76 ...... R
GW Bush_1 ...... 0.99 ...... R
Nixon_2/Ford ...... 1.02 ...... R
Trump ...... 1.12 ...... R
Nixon_1 ...... 1.39 ...... R
Reagan_1 ...... 1.54 ...... R
Obama_2 ...... 1.64 ...... D
GHW_Bush ...... 1.72 ...... R
Clinton_2 ...... 1.74 ...... D
Carter ...... 1.74 ...... D
Reagan_2 ...... 1.86 ...... R
Clinton_1 ...... 1.92 ...... D
Obama_1 ...... 2.08 ...... D

Note that all of the bottom 6 are R, while a majority of the top 6 are D. Doing a t-test for difference of means gives a p-value around 0.01, which most people would categorize as highly statistically significant.

I was surprised by this - the sample size is relatively small and there's a fair amount of overlap. But the mean difference is clear and apparently significant.

Interestingly, the current crash of the stock market doesn't have all that much effect on the result. Using the pre-crash peak (Feb 19, 2020) doesn't lift Trump out of the bottom 6 and doesn't change the finding of statistical significance (p-value is still below 0.03).

So ... contra David in Cal, the sample size is adequate to demonstrate statistical significance, using the full data set. Over the past 50 years, the stock market does better under Democratic presidents.

Minor caveats: Nixon's first term and Trump's term are incomplete. Ford is lumped in with Nixon's second term.

Ned said...

Forgot to provide the actual values for the means:

Democrats: +82% per term
Republicans: +35% per term (using pre-crash peak for Trump)

David Appell said...

Ned, thanks for that.

Here's another stat: Trump likes to brag about the economy and jobs. But (and this is also pre-crash):

job gains in Trump's first 37 months: 6.92 M
job gains in Obama's last 37 months: 8.25 M

Granted, job gains may be more difficult when the economy is near full employment....

David in Cal said...

if Trump gets credit for confidence instilled immediately after the Nov 2016 election, why doesn't Obama get credit for the long recovery after the 2008 financial crisis?

David, I think I have already provided my answer here, although you may not agree with it. Two big reasons:

1. Timing. After a recession, it's normal to have an economic recovery like the one during Obama's Presidency. OTOH It was unheard of to have a continued economic expansion after a recovery of that duration.

2. Policies. Obama had few policies that could be expected to benefit the economy, and several that could be expected to hurt the economy. OTOH Trump had a substantial number of significant policies that could be expected to boost the economy.

Cheers

Ned said...

Or perhaps it's the other way around: "Obama had many policies that could be expected to benefit the economy, and few that could be expected to hurt the economy. OTOH Trump had a substantial number of significant policies that could be expected to hurt the economy."

Best thing to do is stop engaging in special pleading and just go with the data.

There are 13 presidential terms with data in FRED (5 DEM, 8 REP). The DEM terms have a mean increase of 82%. The REP terms have a mean increase of 35% (or less with the current crash).

The difference is statistically significant (p-value approx 0.02 depending on methods).

Those are the facts. Of course it's perfectly reasonable to say you prefer to vote REP in spite of the fact that they're worse for the economy (e.g., if you support a right-wing social agenda).

Layzej said...

What about American Recovery and Reinvestment Act of 2009? $831 billion was injected into the economy, which didn't recover by chance.

Ned said...

Yes, ARRA was of course critical, but in an ideal world it would have been significantly larger. At the time the data on the depth of the 2008 economic collapse weren't complete, and the Obama Administration somewhat lowballed the stimulus program, in part due to sincere beliefs and in part to appease Republicans and conservative Democrats.

This is an issue we could face again as early as next year. Past history shows that Republicans will run up the debt when they are in power, and then turn on a dime and demand steep budget cuts when a Democrat wins the White House. That continual gaming of the system needs to stop.

David in Cal said...

Ned - People very often apply statistical formulas where they don't actually satisfy the necessary conditions. The data you quote mean nothing. The p-value means nothing. These Presidential terms are very far indeed from being samples of the same, exact distribution. A given party's Presidential terms are not all the same. Each one is unique. For that reason, these statistical arguments are invalid.

Cheers

Ned said...

More special pleading.

Of course each presidential term is not the same. So what?

Each individual person is not the same - does that mean you can't use a t-test to establish that the mean heights for men and women are statistically significantly different? Each person is unique!

Likewise, assume that the USA lasts for 4000 years, and we have data on 1000 presidential terms. Does the fact that "each one is unique!" mean we can't use a statistical test to determine whether a difference in mean growth between two parties' presidents is significant?

What about 10,000 presidential terms? They're still all unique........ still impossible to analyze them statistically?

FYI, one of the primary advantages of the t-test is its suitability for small sample sizes...

nowadaysclancycantevensing said...

The Obama Administration wrote the Sept 2010 thru Sept 2017 yearly Federal Budgets yet DavidinC claims Trump was responsible for the market rise of Nov'16 - Jan'17 before he was even Prez. LOL.

Prior to Sept 2009, and prior to the huge WS bailout which happened after GWBush signed into law legislation meant to stop Enron type collapses which W promised would never happen again even as he worked behind the scenes to undermine that law, Republican Administrations who wrote the Federal Budget were responsible for 78% of the nearly $12 Trillion Nat debt in Sept 2009.

78%

Obama = DJI rise during his admin, 300%
Trump = DJI " " " " so far, 35%

Political hacks will of course poo poo it all b/c a Con is running the show now.

Layzej said...

Yes, ARRA was of course critical, but in an ideal world it would have been significantly larger.

It seems to have given the economy the jolt it needed. Having conservatives to push back probably landed the USA in the sweet spot between spending enough and over spending.

Canada's stimulus package in 2009 was about half as much per capita. This was under a conservative government. Our current liberal government is already running a deficit, which compromises our ability to react to the current crisis. The promise of 'modest' deficit spending is partly why they won the election in 2016. We may regret our choice there.