Wednesday, 7 October 2015

Close To The Edge

Once again, I am indebted to my old friend Scott for drawing my attention to some breaking news - a sports betting scandal over in the USA.

Unlike the Betfair scandal, which sees UK regulators either incapable or unwilling to address, it's unlikely to take the US authorities long to act, unless guns are involved in which case it's business as usual. Got mental issues? Sure, no problem Sir - how many guns would you like?  13? - unlucky for some, here you go

Scott notes that;
According to this article, US federal law accepts that fantasy leagues where you bet on the results produced by individual players constitutes a game of skill rather than chance.

But betting on match outcomes themselves is somehow different.

Totally bizarre.

Legalising sports betting is the only (obvious) solution to the problem.
Equally as bizarre is that internet poker was considered a game "subject to chance" and banned in 2006. Buying stocks, options, futures etc. is all skill then. Right.

Here is the New York Times article on the US Daily Fantasy Sports scandal:
A major scandal is erupting in the multibillion-dollar industry of fantasy sports, the online and unregulated business in which players assemble their fantasy teams with real athletes. On Monday, the two major fantasy companies were forced to release statements defending their businesses’ integrity after what amounted to allegations of insider trading, that employees were placing bets using information not generally available to the public.
The statements were released after an employee at DraftKings, one of the two major companies, admitted last week to inadvertently releasing data before the start of the third week of N.F.L. games. The employee, a midlevel content manager, won $350,000 at a rival site, FanDuel, that same week.
“It is absolutely akin to insider trading,” said Daniel Wallach, a sports and gambling lawyer at Becker and Poliakoff in Fort Lauderdale, Fla. “It gives that person a distinct edge in a contest.”
The episode has raised questions about who at daily fantasy companies has access to valuable data, such as which players a majority of the money is being bet on; how it is protected; and whether the industry can — or wants — to police itself.
The leagues have been swelling in popularity, their advertisements blanketing football game broadcasts.
The industry has its roots in informal fantasy games that began years ago with groups of fans playing against one another for fun over the course of a season. They assembled hypothetical teams and scored points based on how players did in actual games.
But in recent years, companies, led by DraftKings and FanDuel, have set up online daily and weekly games based on a similar concept in which fans pay an entry fee to a website — from 25 cents to $1,000 — to play dozens if not hundreds of opponents, with prize pools that can pay $2 million to the winner. Critics have complained that the setup is hardly different from Las Vegas-style gambling that is normally banned in the sports world.
On Monday, DraftKings and FanDuel released a joint statement that said “nothing is more important” than the “integrity of the games we offer,” but offered few specifics about how they keep contests on the level.
A spokesman for DraftKings acknowledged that employees of both companies had won big jackpots playing at other daily fantasy sites. Late Monday, the two companies temporarily barred their employees from playing games or taking part in tournaments at any other site; they already had prohibited their employees from playing on their own company sites.
“Both companies have strong policies in place to ensure that employees do not misuse any information at their disposal and strictly limit access to company data to only those employees who require it to do their jobs,” the statement said. “Employees with access to this data are rigorously monitored by internal fraud control teams, and we have no evidence that anyone has misused it.”
Industry analysts said the episode could leave the leagues open to further criticism that they are too loosely regulated.
“The single greatest threat to the daily fantasy sports industry is the misuse of insider information,” Mr. Wallach said. “It could imperil this nascent industry unless real, immediate and meaningful safeguards are put in place. If the industry is unwilling to undertake these reforms voluntarily, it will be imposed on them involuntarily as part of a regulatory framework.”
Already, there has been intensifying discussion on social media and among lawmakers over whether daily fantasy games are pushing the boundaries of an exemption in a 2006 federal law that has allowed them to operate. The law prohibited games like online poker but permitted fantasy play, deemed games of skill and not chance, under lobbying from professional sports leagues. The games are legal in all but five states.
But because Congress did not foresee how fantasy sports would explode, one member, Representative Frank Pallone Jr., Democrat of New Jersey, recently requested a hearing to explore the relationship between fantasy sports and gambling. “I really think if they had to justify themselves at a hearing they wouldn’t be able to,” Mr. Pallone said in a recent interview.
The data that DraftKings acknowledged was released by its employee, Ethan Haskell, showed which particular players were most used in all lineups submitted to the site’s Millionaire Maker contests. Usually, that data is not released until the lineups for all games are finalized. Getting it early, however, is of great advantage in making tactical decisions, especially when an entrant’s opponents do not have the information at all.
A spokeswoman for DraftKings said that Mr. Haskell simply made a mistake and that the company was certain he did not use the information improperly. She declined to go into specifics about the safeguards or the company’s auditing policies.
Representatives of both companies acknowledged that many employees of daily fantasy companies were players first and had continued to compete on other sites. Ben Brown, a founder of Daily Fantasy Sports Report, was first to disclose that Mr. Haskell had posted the information. Mr. Brown also said a FanDuel employee with access to its internal data, Matthew Boccio, had played on DraftKings; a FanDuel spokeswoman confirmed that.

"There’s a significant amount of crossover,” said Chris Grove, an industry analyst and editor of “The nature of the industry is so specialized and so new that, at the speed which they grew, they relied heavily on the player population.”
Many of these employees set the prices of players and the algorithms for scoring. In short, they make the market.
As daily fantasy sports has blossomed into a multibillion-dollar industry in the past year, DraftKings and FanDuel have become cherished sponsors of M.L.B. and N.F.L. franchises.
Eilers Research, which studies the industry, estimates that daily games will generate around $2.6 billion in entry fees this year and grow 41 percent annually, reaching $14.4 billion in 2020.
So high are the potential financial rewards that DraftKings and FanDuel have found eager partners in N.F.L. teams, even as league executives remain staunch opponents of sports betting.
Jerry Jones of the Dallas Cowboys and Robert K. Kraft of the New England Patriots have stakes in DraftKings, which recently struck a three-year deal with the N.F.L. to become a partner of the league’s International Series in Britain, where sports betting is legal. In addition, DraftKings has tapped hundreds of millions of dollars from Fox Sports, and FanDuel has raised similar amounts from investors like Comcast, NBC and KKR.
Adam Krejcik, a managing director at Eilers Research, said early missteps were often part of booming growth in a new and often misunderstood sector like daily fantasy sports. He said whether Mr. Haskell, the DraftKings employee, made an innocent mistake or not, the damage was done.
“Certainly does not look good from an optics standpoint, and it strengthens the case for additional oversight and regulation,” he said.
Mr. Grove, of, said this may be a watershed moment for a sector that has resisted regulation but now may need it to prove its legitimacy.
“You have information that is valuable and should be tightly restricted,” Mr. Grove said. “There are people outside of the company that place value on that information. Is there any internal controls? Any audit process? The inability of the industry to produce a clear and compelling answer to these questions to anyone’s satisfaction is why it needs to be regulated.”

Tuesday, 6 October 2015

Curse Of The Billy Goat

October became a little greener on the final day of the regular MLB season with both the Los Angeles Dodgers (~1.29) and the Pittsburgh Pirates (~1.48) both winning. That makes it seven from seven for this season and over five years since the last sub 1.51 hot favourite lost an October regular season game.

The third profitable season in four, and the record of this strategy over the last twelve seasons looks like this:

Over the twelve seasons, the strategy is most profitable in April (+22.55) and least profitable in September (-64.90) but clearly things have changed and the conclusion of a famous study from 1994, updated in 2003, doesn't seem to hold true today*.
In contrast to the consistently observed favourite–longshot bias found in racetrack betting markets, it has been shown that gamblers in the market for Major League Baseball games reveal the opposite behaviour. This paper updates the previous study with ten years of additional data for the 1990–99 seasons. The strength of the reverse favourite–longshot bias is virtually identical to the original paper. The result suggests that, contrary to most reported inefficiencies in gambling markets, this bias appears to be permanent.
Permanent? Never say never. The strategy is profitable from as far back as 2008, with the most recent eight seasons +34.06. Since the 2007 season, every month with a full schedule is profitable bar May, which strangely was the most profitable month when the reverse  favourite–longshot bias was, while waning, still present.

It's also interesting that fewer teams, a lot fewer in fact, have been starting at 1.5 or less in the last few seasons. The 356 in 2004 became just 124 ten years later. Baseball punters becoming a little sharper?

Anyway,regular season is in the books now, and by the time the new season rolls around next April 4th you'll have all forgotten this, so it's on to the play-offs which start tonight with the New York Yankees visiting Houston for a one game play-off. The Astros are slight favourites, and the following evening the National League's Wild-Card play-off game sees the Chicago Cubs favourites to win in Pittsburgh.

While laying all favourites has been profitable in the play-offs since at least 2004, (only in 2009 would more than a small loss have been the outcome), laying away favourites has been more profitable than  home:
Should the strategy prove successful in Pittsburgh on Wednesday night, it will of course mean the Curse of the Billy Goat will continue for at least another year. It's only been 107 years - what's another one?

* Another study noted that:
Woodland and Woodland (1994) argued that betting in baseball yielded a reverse favorite-underdog bias, with underdogs underbet. Gandar et al. (2002) made a minor correction to the Woodland-Woodland methodology and found that if there were any bias, it was very slight.

Sunday, 4 October 2015

Green October

The 100% October record for sub 1.51 favourites continued yesterday with the Los Angeles Dodgers beating the San Diego Padres by a single run. The Pittsburgh Pirates drifted to ~1.53 and ended up not being a qualifying selection, which was a good thing because they lost.

It's déjà vu all over again, as the recently departed Yogi Berra once said. Today our favourite pitcher is in action again, and even though the Dodgers have nothing to play for having already won the NL West, they are around 1.32 with Clayton Kershaw to beat the Padres again today. 1.32 will be the second shortest price of the season. There's not much chance of this price drifting past 1.50, but the Pittsburgh Pirates are possible qualifiers currently around the 1.47 mark. This one could drift, but either way, this will be the end of betting on hot favourites in baseball for a few months. The play-offs are a completely different ball game, pun intended.

Hopefully some of you have made some money on this strategy since I made the observation just before the All-Star Break in early July. Since that day, there have been 84 winners from 115 selections with a profit of 15.95 points - an ROI% of  13.9%. 

Saturday, 3 October 2015

Regular Season October Baseball

It's the final weekend of the 2015 MLB regular season, and a good time to look at whether or not it is a profitable strategy to back the hot favourites in the play-offs. Unsurprisingly there are fewer such opportunities in post season play, an average of 10 a season if we define hot favourites as those with an implied probability of 60% or greater (1.6).

If we relax the constraint on 'hot' to 'tepid' 55% or greater implied probability, we get more selections - but an even worse outcome.
And if anything, the returns on favourites in the play-offs have declined over recent seasons.
The competitive nature of the post-season appears not to be kind to favourites. Note that down for the next few weeks.

Jeffrey's observations on the fate of favourites and underdogs at different times of the season resulted in me taking at look at results by moth over the last five seasons. Only May shows a loss for hot favourites (sub 1 .51) but I'll wait for the 2015 regular season to finish this weekend before publishing all the numbers.

With a couple of likely qualifiers today, namely the Los Angeles Dodgers and the Pittsburgh Pirates, some of you may be interested to know that hot favourites in October have a 100% record at 6-0 since 2011. Bizarrely, three of those six were on one single day, which was yesterday, and all three won. Happy days.
The Dodgers tonight are currently 1.41 on the exchanges, with the Pirates at 1.48. 

Hopefully, it's not meaningful that the six winners have all been against Right-handed starters and tonight's are both Lefties!

Friday, 2 October 2015

Pre v Post All-Star Break

Jeffrey had a comment on my Studies Have Continued post saying:

I'm on the other side of this coin, backing dogs coming off a win in certain situations.
One thing I've noticed for three years running now, is that it all goes south in the month after the All Star break. Suddenly favourites predominate until around mid-August onwards, when my dogs start winning their share again.
I can only think that by July-August, the bookies have got the measure of the teams when pricing games up, and it takes the back-end of the regular season, when teams gather themselves for one last tilt at a play-off spot (or in readiness for off-season contract negotiations!) that certain underdogs start over-achieving again.
I just wondered if you had noticed this trend and, if so, if you had any theories as to what might cause it.
I like these thought-provoking comments and did a little research. I took a look at how shorties (1.50 or shorter) perform in the period leading up to, and following on from, the All-Star break which is held somewhere around the second week of July. I compared June and pre-All-Star July with post-All-Star July and August for the past five seasons.

The results showed that over that period, backing shorties both before and after the All-Star break was a profitable strategy, with the 'after' period showing a slightly more profitable ROI%. while the win percentage was almost identical. With the All-Star game usually in the first half of July, there are more qualifiers in the 'after' period.
Over the past three years, Jeffrey's period of observation, late July / August showed a profit of 28.80 points after versus a loss of  5.65 points before the break when backing all shorties.

Without knowing what the 'certain situations' are that determine Jeffrey's selections, or the price ranges of his underdogs, it's impossible to say much more, but hot favourites have been pretty solid recently in the 6  / 7 weeks following the All-Star game.

Jeffrey suggests that "by July-August, the bookies have got the measure of the teams when pricing games up" but the days when bookies priced games up are long gone. Prices are determined by punters, and we are collectively pretty accurate. Sportsbooks such as Pinnacle Sports and the exchanges lead the way of course, with the more traditional bookies shadowing those prices and allowing anyone with half-a-clue to bet a whopping 25p if they're lucky.

Thursday, 1 October 2015

Wanted - Active Investors

At the end of a rare second consecutive losing month for stocks, it was interesting to read the comments of Goldman Sach's Tim O'Neill on the topic of active versus passive investing, although given that he is 'primarily in the business of active investing', his opinion on the subject is not the biggest of surprises:

Tim O'Neill, Goldman Sachs' partner and global co-head of the investment management division, has a warning: If passive investing gets too big, then the market won't work.

"So in terms of the size, a market needs both active and passive investing because if everybody's a passive investor, there's no one to buy from," O'Neill said on a new "Exchanges at Goldman Sachs" podcast with communications chief Jake Siewert. "And if passive becomes a certain oversized percentage of the market, the market doesn't function."
O'Neill is getting to the heart of the debate on active vs. passive investing. Passive investing has boomed in recent years, with index-tracking exchange-traded funds hoovering up trillions in assets under management.
The problem is that the market needs active management, otherwise there will be no one to buy from, and individual stocks will just move with the overall index.
As a result, O'Neill, who previously called passive investing "a potential bubble machine," said that both strategies are necessary.
Here's an excerpt:
Well, the promise of active investing is that they're going to deliver performance net of fees better than the benchmark, whatever the benchmark might be for the US, global, or Europe. It's been a difficult seven years for active investors because the markets have risen so consistently and persistently higher. So most active managers have net of fees underperformed the benchmark. So it led back to this debate about whether or not the fees that you pay for active managers are worth it. And there's been simultaneously a great shift towards passive investing because it's cheap, and you would get all of the market returns, net of five or 10 basis points.
The problem for passive is that its size, at a certain point, may be too much for the market to handle. And it's also all on autopilot. So in terms of the size, a market needs both active and passive investing because if everybody's a passive investor, there's no one to buy from. So there's no one ... your beta is my alpha and vice versa. So you need a balance in the market. And if passive becomes a certain oversized percentage of the market, the market doesn't function.
The other problem with passive, of course, it's all on autopilot. And when you get to periods of misvaluation, over or undervaluation, you need active decision-makers. Because valuation always matters in markets, and investing.
Peter Andersen of Forbes has a slightly different opinion:
Rather than choose one of these philosophies over the other, I take the unpopular stance of embracing them both, and there are very good reasons to do so. Let me explain how I came to this position. Take a look at the (above) graph, which shows schematically what percentage of active managers outperformed the passive S&P 500 index investor.
Clearly there is some type of vague cyclicality here. There seems to be a horizontal line at 45% around which the graph oscillates. At some points such as March 2005 to March 2006, and again in March 2009 to March 2010, a high percentage of managers outperform the S&P. On the other hand, these periods of strong out-performance are followed by sudden drops in performance. That is, the passive investor beat the active stock picker. I’m proposing the following behavioral reasons for the cyclicality:
Picture many highly skilled analysts all competing to obtain and synthesize relevant information on a stock before everyone else. So many are working so hard, that it is difficult to uncover valuable information before others. At the extreme point, where all information is known by all hard-working analysts, there is virtually no advantage to staying in the game if you’re hoping to gain an information edge. Ironically, staying–and expecting not to gain an information edge–would ensure that no one else gains the information either.
If all keep working as hard as possible and none drop out, active management has no edge. With no information edge, indexing begins to look like an appealing alternative and may very well be outperforming active management. But what if some participants become discouraged from the lack of return on their efforts, and they drop out of the active circle and choose to index? That creates the opportunity for the other active die-hards that haven’t given up.
I propose that this dynamic is behind the cyclical shifts of active versus passive performance:
When market participants become frustrated by the lack of out-performance of active management, some exit the active arena, choosing instead to index. That very exit from the active arena sets the stage for the remaining active managers to outperform. The siren song of active out-performance then lures those participants back in the game. But when everyone piles into active management, the ability to gain an information advantage diminishes, causing the cycle of switching back to passive again. And thus the cycle continues. In an odd way, one could argue that the oscillations between active and passive promotes a type of market stability.
Examine the graph again, this time keeping my theory in mind. I’m sure you will see the pattern makes more sense. When considering the choice of active versus passive, a more reasonable answer is to open your mind to both alternatives, not just one.

Friday, 25 September 2015

Studies Have Continued

In my post "Studies Have Shown" from July, I pointed out that the strategy of backing baseball shorties has become profitable in recent seasons, after years of this sport being the poster-boy for the reverse favourite-longshot bias.

At the time of that last post, 2015's profit was a solid 7.45 points, and this has since climbed to 24.00 points with the subsequent 90 qualifiers generating another 16.55 points.

A more complete look at this strategy going back 12 seasons is below, and I'll update this again at the end of the season.
Some of you may have seen that this blog's most closely watched pitcher Clayton Kershaw, took his first loss since June at the weekend, although he returned to winning ways last night. Opposing Kershaw since I first mentioned the idea would have made 5.48 points (no prices yet from last night's game):
That 1.28 on 14th September was only the second sub 1.30 price this season (the other being Kershaw's 1.27 on July 8th). There was also only one such price last (2014) season, and that was also a Kershaw game.

Some of you may also have noticed the trend towards Unders on Kershaw's games - from 30 starts this season, the profit would have been 6.40 points. Impressive, and over Kershaw's career, backing Unders would be up 20.15 points (246 games). Of course Kershaw was an unknown as a rookie, so no one on Earth would have actually done this, but since the 2010 season when he did become rather well known, the Kershaw / Unders record has been solid:
A change in personal circumstances has meant that I have had to scale down my blogging, and indeed betting, activities as some of you may have noticed, but I do still try to keep up with some of those blogs on my blog roll, and found this entry from my old friend Geoff:
The King of Bloggers! As the saying goes, Once a King, always a King, but once a Knight is enough.  

stats counter
page counter