By Malay Bansal

Much has been said and written about the surprising results of Brexit vote in UK and Trump win in US Elections, their causes, and implications. However, there is an obvious and important lesson for those managing investments that has not received much attention.

Note: A version of this article was also published on TalkMarkets.

The unexpected wins for Brexit in the June EU referendum in UK, and for Donald Trump and the republicans in November elections in US, and what these results mean for the economy and markets, have undergone a lot of discussions. These are very important and relevant topics. However, the implications of those wins are not the focus of this article. Rather, the focus is on examining the events from the perspective of investment managers, whose investment decisions are shaped by views formed from the information, opinions, and forecasts they receive from the media and the experts. The focus here is on the fact that in two of probably the most consequential events of the year, almost the entire world reached the wrong conclusion, and yet there was readily available information that would have pointed anyone, who took a slightly deeper look at the data rather than simply accepting the wide consensus view, to a better conclusion, or at least, not as certain a wrong conclusion.

BREXIT Vote

On 23 June 2016, the United Kingdom voters defied the polling experts, betting odds and much of the political establishment, and surprised the world by voting 52% to 48% in favor of leaving the European Union. Most people had gone to sleep believing Remain win was a foregone conclusion. The surprise roiled the markets – within an hour of the announcement of the result, the Pound lost 10% of its value, US Treasury yields were down 30-35 basis points, and Gold price was up by $73.

Before the vote, the opinions polls were close, with Remain slightly ahead[i].

brexit-opinion-polls

With lack of clarity from polling results that were veering all over the place, markets were focused on probabilities derived from betting markets. Betting odds are assumed to convey the “wisdom of the crowds” and take into account a greater range of factors than the snapshot a poll will offer. The evening of the referendum, bookmakers (including U.K. bookmaker William Hill PLC and British bookmaker Paddy Power, part of Betfair Group PLC) placed the chances of a “Leave” vote at 10% or less. By Thursday night as the polls closed, Betfair was predicting a 94% chance the U.K. would vote to stay, and the pound reached its highest levels of the year against the dollar.

eu-referendum-betfair

There was information out there from bookmakers that would have questioned the conclusion based on odds from them. But it seems like almost no one was looking for them as the opinion had been formed and consensus reached already.

Reports on June 3[ii], quoted spokesperson for the betting shop Ladbrokes saying that over 80% of the number of bets outside London were in favor of Brexit, even though Ladbrokes’ odds were in favor of Remain. Similarly, British bookmaker Paddy Power, part of Betfair Group PLC, said its odds skewed in favor of a vote to remain due to large bets. Graham Sharpe, a spokesman for William Hill, said 71% of bets placed with the firm are on a “Brexit — British exit — outcome, but 73% of all the money bet with the company is on the remain camp prevailing

Ladbrokes Bets.jpg

Clearly fewer larger bets were skewing the odds in favor of Remain even though majority of bets were for Exit. Unlike a market were the size of purchases is important, in a poll or referendum, one person gets just one vote. Ladbroke publicly talked about the data before the vote and after the vote explained why the betting markets got EU Referendum results so wrong[iii] in a blog. If most of the cash went on Remain, as it did, bookies had to follow the money and make Remain the favorites. And the bookies all agreed that while three-quarters of the £40 million (not a very big amount in the scheme of things) eventually gambled on the referendum was placed on Remain, when it came to counting individual bettors, bets on Leave far outnumbered punts on staying in the EU.

Very few market observers focused on this data (“Something Strange Emerges When Looking Behind The “Brexit” Bookie Odds[iv]) before the vote, even though the stakes were high. People knew big market moves were likely if Leave side won. One hedge fund manager reportedly made £110 million[v] by betting Britain would vote for Brexit and the pound would tumble. George Soros correctly predicted[vi] possible 15% decline in value of Pound if Brexit won, but even he is reported to have had a long bet on Pound.

Trump Victory in US Elections

As polls were coming to a close on November 8, the prevailing view in the media was a victory for Hillary Clinton and a loss for Republicans. As counting began, within a few hours, it started becoming clear that all of the polls, predictions, and political pundits’ views were simply wrong. Just as in the case of Brexit vote, most everyone had reached the wrong conclusion. The surprise was evident in markets – the Dow Jones Industrial Index futures went into a freefall around 8 PM, quickly falling 900 points or about 4%. What is remarkable is that the forecasters got it wrong even after the Brexit experience and with so many comparing US situation to Britain!

Just as in case of Brexit vote, people have analyzed the causes for the surprise result – the socio-economic factors, the failure of polling techniques, etc. Much has been said and written since then (“5 surprising lessons from Trump’s astonishing win[vii], “Five key lessons from Donald Trump’s surprising victory[viii], “Election Experts Puzzled Over Surprise Trump Victory[ix], “Trump tells Wisconsin: Victory was a surprise[x]).

Before the election results were out, the widely-followed FiveThirtyEight forecast[xi] gave a 71.4% chance to Hillary win, TheUpshot in New York Times gave her an 84% chance[xii], and betting site Paddy Power gave her a 4/11 odds or 73.3% chance[xiii].

The actual polls were much closer. NYTimes.com summarized various polls[xiv]:

ny-times-polls-summary

On the face of it, though the difference was only 3 points, almost 9 out of 10 polls indicated a Clinton win. However, if you looked at the details, that could lead you to question that as a foregone conclusion. For example, below is the result of the Bloomberg poll results published on Nov 7. Clinton was ahead of Trump by 3%, but interestingly 4% were listed as “Don’t want to tell.” If one thought about whether Hillary or Trump supporters were more likely to hide their support, and noted the +-3.5% margin of error, the 70-80% probability of Hillary win might have seemed questionable, since even though 9 out of 10 polls showed her winning, each one was highly uncertain. Another important factor that many people (though not the professional forecasters) seemed to have not paid attention to is the fact that what matters for who wins the presidency is the outcome of electoral college voting not just the popular vote (Hillary Clinton did win the popular vote marginally but did not win Presidency). So, again, the jump to conclusion by majority that Hillary Clinton was going to win the presidency was not justified based on information that was available.

bloomberg-poll

Just like Brexit, there were signs that the majority did not pay attention to: “The clearest sign yet that the US election will be like Brexit[xv], “Why Brexit could be a warning for American voters, Trump and Clinton[xvi].

These two cases of a view held by the vast majority being wrong is not an isolated phenomenon either. And they do not always get corrected quickly. The Great Credit Crisis was caused by the almost universal belief that US house prices could not go down or could not go down everywhere at the same time! The huge successes of the few who questioned that belief[xxii] and did their own work have been chronicled in article, books, and movies.

The incorrect views are also not necessarily always those ignoring risks. In 2008 and 2009, majority of investors in mortgage area had the view that commercial mortgage sector was going to be similar to subprime and AAA CMBS were going to take huge losses. Unlike subprime AAAs, CMBS AAAs have not taken losses and are trading above Par, but it was very hard to convince people to buy those bonds when they were trading at 50-60% of Par (I was in the process of raising a fund to buy AAA CMBS bonds at the time).

Lessons for Investors from Brexit Vote and Trump Victory  

The almost universal consensus views were wrong in both of these major events. The data that did not support the conclusion was there and available – just one google search away. Yet, it seems like most people did not look for it, or pay attention to it. The Efficient Market Hypothesis tells us that the market prices always incorporate all available information. Yet, despite the free, easy, and abundant availability of information to everyone, that was not true in both of these highly consequential events that had the attention of almost everyone. What the market really reflects is interpretation or views of market participants, which can clearly be wrong at times.  One obvious point for investment managers to keep in mind is that those who dig into details, look for relevant data, and draw their own conclusions can still add significant value to the investment process.

It is also useful to examine why were the beliefs so universal and so one-sided? Several factors played into it. How information flows from the source to its consumers is one of the factors[1]. One feature of that flow of information today is that it gets repeated several times. Once a statement is made, an announcement is released to press, a blog is published, or a tweet is sent, it gets picked up by multiple sources, repeated on other news channels, republished on other blogs, or retweeted. One somewhat comic example that illustrates the point is a clip showing news anchors from different TV stations all repeating the same phrase[xvii]Consumers of information hear and see the same thing from multiple sources, even though it all may have originated at a single point. Reading and hearing the same thing from multiple sources makes it more believable. The more people start believing it, the more they repeat it[2], and pretty soon everyone has the same view, and people stop questioning it or even looking for data that might show a different conclusion.

We get a lot of information in our inbox, a lot of which is often repetitive. Spending time on those gives us false confidence that we have a really good handle on the situation. However, in reality that leaves less time for digging into details beyond the headlines and the confident view is not based on full information. In some ways, getting too much information may be harmful rather than helpful.

For some time, the economy has benefited from the recent technological advances in internet and web space, and the resulting free information available to all, which has contributed to higher productivity. But now, that phenomenon may have reached the stage of diminishing or negative marginal returns. Too much information coming in tends to cause people to focus just on the headlines and less on digging in to details, or questioning something that looks like everyone in the world agrees on. A seeming consensus from everyone makes people less likely to spend time questioning it. Even though the information exists and is easily and freely available, we may not make any attempt to look for it because we believe we already have it. Those who ask if the consensus is an informed view based on full analysis will look for more data.

Another factor is blurring of lines between journalism and entertainment/editorial/commentary. News-type shows often are influenced by personal or corporate views of those presenting the news, or are highly correlated to their own biases. There is also a blurring between facts, opinions and conclusions, with all often presented or perceived as facts.[3]  Combine those with the fact that most people prefer certainty (easier to think about) over uncertainty, and it is easy to see how a wrong consensus view is possible.

While looking out for biases in others, we also need to be aware of our own biases. With all the choices of information, we chose who we listen to and who we ignore. On social-media, which is becoming a larger source of news consumed, we are linked to people who are generally of same views as us.

A more recent phenomenon adding to the negative marginal returns from information availability is the rise of fake news stories that are put out and spread by people as pranks, or to generate advertisement revenue, or to promote their own agendas by misleading others. These can be inconsequential to serious at times. One recent example[xviii] is the nuclear threat against Israel by the Pakistani Defense Minister[xix] after reading a fake news story. Another incident[xx] involved firing a gun in a Washington pizzeria by someone after reading a false claim online[xxi] about that pizzeria.

From a perspective of longer-term investment opportunities, my view is that we may be reaching a point where the pendulum will slowly start swinging back from the zillions of free sites to paid sources of information, where people can be sure that the information has been vetted before reporting. Maybe the large traditional news organizations will have a more profitable future after all.

My most important take-away  is a simple reminder to ourselves, as investment managers for our own portfolio or for clients’ portfolios, that despite what Efficient Markets Hypothesis tells us, we can add significant value by thinking independently, questioning consensus views, working intelligently (making an effort to look for details beyond just the headlines), and considering biases – in our sources of the information, and our own.

Note: The views expressed are solely and strictly my own and not of any current or past employers, colleagues, or affiliated organizations. My writings are simply expressions of my intellectual thought process. The intent is not to promote any particular view point or agenda, and the writings are not influenced by any other groups or individual.

 

 

[1] I have used the term Information Momentum to describe the concept. See Using Information Momentum to Understand Markets & Economy (https://marketsandeconomy.wordpress.com/2014/08/03/using-information-momentum-to-understand-markets-economy/).  Everyone seems to form the same opinion at the same time. However, sometimes those opinions can be wrong.

[2] Web and internet make it very easy for anyone to repeat and reach a very wide audience.

[3] Howard marks makes some great points in the Latest memo from Howard Marks: Expert Opinion (https://www.oaktreecapital.com/insights/howard-marks-memos).

[i] https://today.yougov.com/news/2016/06/22/final-eve-vote-brexit-poll-remain-leads-two/

[ii] http://www.businessinsider.com/most-gamblers-think-brexit-will-happen-but-odds-still-favour-remain-2016-6

[iii] http://news.ladbrokes.com/politics/british-politics/why-did-betting-markets-get-the-eu-referendum-result-so-wrong.html

[iv] http://www.zerohedge.com/news/2016-06-22/something-strange-emerges-when-looking-behind-brexit-bookie-odds

[v] http://www.dailymail.co.uk/news/article-3666800/Harrow-educated-hedge-fund-manager-won-110MILLION-betting-Britain-vote-Brexit-pound-tumble.html

[vi] http://fortune.com/2016/06/20/soros-brexit-pound/

[vii] http://www.cnn.com/2016/11/09/politics/donald-trump-wins-biggest-surprises/

[viii] https://www.washingtonpost.com/news/monkey-cage/wp/2016/11/09/five-key-lessons-from-donald-trumps-surprising-victory/?utm_term=.5b56c825a927

[ix] http://www.voanews.com/a/election-experts-puzzled-over-donald-trump-surprise-victory/3589558.html

[x] http://www.politico.com/story/2016/12/donald-trump-wisconsin-232605

[xi] https://projects.fivethirtyeight.com/2016-election-forecast/

[xii] http://www.nytimes.com/interactive/2016/upshot/presidential-polls-forecast.html

[xiii] http://blog.paddypower.com/politics/2016/11/02/early-payout-has-paddy-sweating-as-trumps-odds-continue-to-improve/

[xiv] http://www.nytimes.com/interactive/2016/us/elections/polls.html

[xv] https://www.indy100.com/article/brexit-donald-trump-us-election-2016-7386096

[xvi] http://www.cnbc.com/2016/09/20/why-brexit-could-be-a-warning-for-american-voters-trump-and-clinton.html

[xvii] https://www.youtube.com/watch?v=PStpvviPgxk

[xviii] http://www.huffingtonpost.com/entry/pakistan-nuclear-israel-fake-news_us_585f2e8ae4b0de3a08f58fb6

[xix] http://www.nytimes.com/2016/12/24/world/asia/pakistan-israel-khawaja-asif-fake-news-nuclear.html

[xx] https://www.washingtonpost.com/opinions/pizzagate-shows-how-fake-news-hurts-real-people/2016/11/25/d9ee0590-b0f9-11e6-840f-e3ebab6bcdd3_story.html?utm_term=.9a7f77e036ca

[xxi] http://www.nytimes.com/2016/12/05/business/media/comet-ping-pong-pizza-shooting-fake-news-consequences.html?_r=0

[xxii] http://nymag.com/daily/intelligencer/2009/11/bad_news_bears_2.html

By Malay Bansal

A simple concept that is useful in understanding some counter-intuitive phenomena in markets and economy, and looking ahead to the future.

For over a decade, I have used a concept that I have called Information Momentum in my thinking and discussions to explain phenomena in markets and economy that otherwise do not seem completely intuitive. Understanding why something has been different from expected in the past also helps in understanding and looking ahead to what might happen in future.

At its core, the concept is simple and parallels the similar concept in physics. Every physical object has a mass, and if it is moving, it has a velocity. The product of the two is known as momentum. The higher the momentum, the greater the impact that object will have. The higher the momentum, the higher the force required to stop the object from moving or to change its direction.

A somewhat similar concept can be applied to information. In this case, consider the number of people who receive a new piece of information, and can act on it, as equivalent to the mass, and how fast people get this new information as equivalent to velocity. The more the number of people who receive the new information, or the quicker the new information reaches people, the higher the information momentum for that information.

The concept is simple and so are some observations which help apply the concept to understanding market behavior. Somewhat in jest, and continuing the parallel with physics, I have sometimes referred to them as my laws of information momentum. Four of these observations are mentioned below.

First law of Information momentum is that the information momentum will continue to increase with time. This is obvious today. It started increasing with with internet becoming more easily available to more people, first in US and then around the world. Then, each of the following, as it came on the scene, has contributed to a quantum jump in the information momentum: advent of the web (the world wide web), email, blogs, news sites, internet trading, faster connection technologies like DSL replacing old phone dial-ups,  wi-fi connections everywhere, tablets & smartphones with data connection, twitter. This trend will continue and the increase in information momentum will magnify the impact of the other laws.

The second law is that higher information momentum means new information will have bigger impact than in the past. More people acting on a piece of information at the same time means bigger moves in market and possibly more often. A corollary to this law is that higher information momentum can, though will not always, increase volatility and correlation in markets.

 The observations by Bespoke Investment Group on “S&P 500 All or Nothing Days” (days where the net daily A/D reading in the S&P 500 exceeds plus or minus 400) as described in the article All or Nothing Days Becoming More Common Than Uncommon, and as shown in the chart below provide a good example of this over a long period.

S&P500 All or Nothing Days Graph

Source: Bespoke Investment Group.

The third law is that more information momentum means better decisions will be made. Better decisions will logically lead to better outcomes, which in bigger picture, implies higher probability of higher profits for companies and better growth for the overall economy. All else being equal, the future will be better than the past. This applies at every level including at the level of individuals, companies, countries, local markets, and the entire world’s economy. At every level, decision makers will have access to more specific and detailed information and sooner than ever before. In addition to more detailed information about the specific situation, decision makers will have access to more ideas, viewpoints, opinions, suggestions, and criticisms from a wide variety of people through blogs, comments etc. As an example, in 2008 and 2009, when the economy worldwide was facing a huge crisis, with declining values of mortgage backed securities and other bad assets  leading the biggest banks towards failure, and the government had to announce extraordinary measures like TARP, even people like me were able to chime in with suggestions directly to people at Treasury and Federal Reserve and via articles in New York Times etc (to toot my own horn, I suggested a plan involving public-private partnership – basically the concept behind TALF and PPIP programs announced and implemented several months later. See Solving The Bad Asset Problem or PDF).

It is easy to see that even with all else being equal, the future will be better than the past. But all else is not equal, if you look at things like what developments like search engines have done to personal and business productivity. You can find answers to almost any question you have just by googling it. Think about how Google Earth has changed the real estate businesses, and other similar examples. All of these are reasons to be optimistic about the future.

The fourth law states that since more information may become available with time, decisions will be made later in time, possibly at the last possible minute when they have to be made. At an earlier time, the Just-in-time concept significantly improved productivity in manufacturing. In a similar vein, decisions to act may be delayed till the last minute so as to take advantage of any additional information that may become available (what I call “Just-in-time decisions”).

Future articles will add more laws and give more specific examples detailing the application of these concepts for understanding various market and economic developments and looking ahead to the future (for the first example of application of these concepts, see Why have U.S. Interest Rates Defied Expectations and What Lies Ahead? ).

Note: The views expressed are solely and strictly my own and not of any current or past employers, colleagues, or affiliated organizations. My writings are simply expressions of my intellectual thought process. I welcome comments, observations, examples and any extensions of the concepts above.

%d bloggers like this: