Topic: Feature Requests / Dividents

I especially like the idea of dividends in very long-term calendar markets, such as the market on when a human will land on the moon again. It is never more than a year away, eventhough almost every participant is aware there are no plans to return until 2020 at the earliest. Also, the per share price is around 500 inkles, scaring away new players. At regular intervals – even annually – the price could be reduced an amount that is paid out to those holding shares. This would keep the market fresh and yield better predictions. We are NOT going to be on the moon in January, 2010, sorry.

Topic: Feature Requests / Timezones

I was wondering if you could add a drop-down menu that allows users to change their time zone (under settings.)

Topic: Feature Requests / Sortable dashboard

Thanks for the post David. Soon, very soon… Definitely at the top of the list of things to do.

Topic: Feature Requests / Sortable dashboard

It would be nice to be able to sort the dashboard output in different ways. The current implementation always goes from most recent activity to oldest, but here are some other views that could be useful.

Grouped by market… If I trade different stocks within a market at different times, they could be scattered all over the dashboard. It would be nice to see them contiguously so I can remember exactly what I own or don’t own.

Sorted by gain/loss… Being able to see my best or worst picks together would let me quickly decide where to take profits or cut losses.

Sorted by close date… Knowing which markets are closing soon would be nice so I can take action before it’s too late.

Any other useful orderings people can think of?

And being able to choose the number of stocks per page would be nice too.

Topic: Feature Requests / Large Trades Distort Prediction Accuracy

Hi Bill
It seems like my point wasn’t clear enough.

In Inkling’s design of a prediction market, the size of a single position can determine the prediction percentage of the total position holders in a market. This means, if a position holder has a large account, they can easily distort the predictive range of agreement, or disagreement.

This isn’t reality in the real world, and it isn’t reflective of crowd estimation research. For example, when I place a trade for 5,000 contracts of Corn, or 3,000 contracts of Crude Oil, if my broker is careful and does it as volume appears in the market, my entry price into the market can end up being near where it was just before I made entry. Now, if other people noticed what I’m doing, and that happens sometimes, they can try to run my stops, or get on in the same direction.

If a lot of people get on in the same direction, this moves the market in the direction of my trade. However, if I’m the only one who gets in, the market will hardly notice my position and the price won’t change much until the volume of trading begins to move either against or in favor of my position. This means that for the market price to change, there needs to be broader market participation, not large size, in order for the price to be significantly impacted.

Inkling’s approach views the size of a large position as broad market participation, instead of a large trader with a big paw. It also means that if you have a large paw, your prediction bias is more likely to be correct than the average of all previous traders. This is a silly concept because it believes “might it right”, through its interpretation of position size. Reality shows that good predictions come from the average results of a crowd, not from the individuals in the crowd who have big pockets.

It isn’t hard to see why Inking might have designed the process to work this way because it simplifies the process of making prediction value adjustments, and because it creates a game that looks like a real world market. However, my argument is that they should have gone in a different direction. For example, let’s assume a large player with 2m Inklings decides to risk 500 contracts in a market, but that position size in comparison to his available Inklings calculates to less than 0.025% of his Inkling equity. Is this large trader thinking his prediction decision is rock solid? It doesn’t seem like it when you compare it to someone with an account of 5,000 Inklings who thinks they know something about the market, so they decide to risk 500 Inklings, or 10% of their account balance in that market. Should the not-so-sure large trader’s position size influence the price more than the person who believes they know what is going to happen, and puts 10%, or a larger piece of themselves at risk? I don’t think it is reality.

My view, if we are to maintain the market model appearance, is to have the risk, in terms of the exposure to a trader’s account be the most influential reflection of his belief be the impacting element when it comes time for Inkling’s Market Maker to adjust the market’s prediction range values. Let’s put this is terms of how each of us would react to two different sets of facts. A new expert who doesn’t know much says, he thinks, “your chances are good, but don’t hold me to a certainty.” However, the next expert says, “controlled research shows that if this approach is used the chance for success is going to be better than 40%, and I’m willing to put my reputation on the line that the information is accurate.”

Would you believe the person who doesn’t want to risk anything more than your frustration, or the person who is willing to take a serious reputation hit, and has supporting data that his answer is correct? My instincts have me going with the person who did the work, stated his certainty and is willing to stand behind what they are saying.

Another example might be when we ask for an opinion and ask a group of people to give us their best guess on what might happen. As we go through the group we ask each person to give us a prediction, and also their belief in how strongly they feel about the certainty of that prediction. If all the people, but one said, their average belief indicates an outcome of a 30% chance for success, but one person demanded we write down the outcome had a 90% certainty and would not stop talking. What would you give the outcome to be from this sampling? Inkling’s calculation would let the stronger person have a large influence on the outcome.

If the Inkling process would look at what people are willing to risk as a percentage of what they have available, then they are getting a percentage number or value from each trader that can then be analyzed to create a prediction percent. This is in reality the information they are getting in the prediction markets that require people use actual cash to describe their belief level. By allowing people to think in terms of how strongly they believe in a prediction, versus the current Inkling paradigm of how much they have to bet, the prediction accuracy would be better and the percentages would move away from what people believe might be a reasonable gamble of where they can make a commitment, to what they think will happen.

Another area that distorts the process is how much the number of Inklings available can allow or restrict how many markets a participant can record their belief in the markets they find interesting. This issue is based in some of the above logic here and because of the market model being presented, but this posting is too long already.

As for the prediction amount being so dependant on large trade size, and allow large traders like yourself to take a large rake from better entry price into a market, I agree that would be a natural result of what I’m proposing here. Of course, that has a downside when the guess is wrong, but that is the beauty of making the process more participation and belief dependant.

When size is less important, accuracy and consistency increase. What we have now are large paws putting low risk positions in a lot of places distorting the reflection of what people think might happen, and doing this with a smaller sample of opinions than would result otherwise. Crowd size is an important aspect for finding the value of something with an opinion process.

Topic: Feature Requests / Large Trades Distort Prediction Accuracy

I’m not sure I understand your point. Of course large trades move the market in a significant way. I don’t understand, however, why that necessarily “distorts” prediction values.

You note that in the real world the impact of a clumsily executed large trade would revert to prior market levels whereas in the Inkling world it doesn’t. In the real world that is only true in the event the large trade is actually distorting (i.e. sets wrong price) as sellers will enter the market and drive down the price to its appropriate equilibrium. If the trade ultimately reflects new information, it will not revert. In either case, there is generally sufficient market liquidity on both sides of the question to set the “right” (one could quibble with that adjective) price, though clearly not in all cases (e.g. the current illiquidity in the mortgage backed markets).

In theory, if a large buy (or sale) in an Inkling market moves a price beyond what is reasonable, other traders are free to sell their positions or short the stock (or buy) and drive it back to where it belongs. Certainly both you and I have done so. Indeed, other traders should thank the distorter for creating opportunities/liquidity to take inklings off the table and/or earn outsized returns for the risk being taken. There are certainly a number of markets with a larger number of traders that appear to be reasonably efficient. If an Inkling market doesn’t revert, presumably other traders believe the the new prediction value is close enough to correct that there is no value to be had in taking a contrary view. In all of the above cases, there is no particular reason to think that the predictive value of one large trader (with conviction) is no better or worse that the predictive value of a bunch of smaller traders. Indeed, if the value predicted by one large trader doesn’t revert in the face of all traders presumably seeing it, the fact that everyone else doesn’t sell arguably implies that the other traders agree so the price does reflect the view of the many.

On the other hand, the alternative is that there is not enough liquidity to drive the price back to where it belongs. However that, presumably, is the distorter’s problem, not anyone elses (OK, maybe the market creator’s if they were using the results for something but they can read in to the lack of trading whatever they want). For what it is worth, I subscribe to that theory – there is not enough liquidity in many Inkling markets.

While I can’t speak to the trading strategies of other large balance traders, as one of them I can only tell you what I do. I almost never invest in more than 50 shares of anything unless I think that a stock is grossly over or under valued based on objective research. Often that means that the market isn’t well followed (no liquidity so minimal predictive value until I enter), or I happen to check in shortly after new markets open with default prices where outsized profits are available. Given that there are relatively easy inkles to be had with that strategy, there is little point in making large uninformed clumsy speculative investments. If anything, I’m guessing that my participation, and that of other large balance players, makes markets more efficient/accurate rather than distorted. In my view, if anything, there are not enough large traders to make every market efficient/accurate – take a look at my separate posts under Locks of the Week – opportunities that should not exist if the masses were paying attention and/or had plenty of liquidity.

Maybe your real complaint is that large balance traders find and exploit the lucrative investment opportunities before the small guys had a chance. Maybe, or maybe the large balance folks are more informed or interested in doing a little research. Last time I checked, the internet was available to every Inkling trader while new markets appear to be launched somewhat randomly. I will admit that large inkle balances probably provide more freedom to take positions in longer pay out markets which is otherwise a sub-optimal investment strategy for lower inkle balance players. I would submit that it makes those markets more predictive though.

To be honest, I’d love your solution to have prices automatically revert to pre-buy levels which would allow me to buy grossly under valued stocks over and over again.

By the way, my potentially “distortive” positions of 250 shares or more are listed below. Please let me know where you think I’ve grossly distorted the true probabilities aside from not buying or shorting enough (i.e. Yes I know the probability of Frei scoring the most points in Euro 2008 is 0, I just don’t have enough inkles to drive it to that level).

1824 long on when the 787 will be delivered
1554 long on the next NASA mission post shuttle retirement
1317 long on when 24 will premier
250 -1,000 shares short in the stocks of primary losers and other non-participants in the New Mexico election markets.
800 short on when the Pittsburgh Casino wil open
600 long on where Boston City Hall will be on 6/21
587 long on when Belgium will have its next astronaut
500 long that a Democrat will have the delegates necessary to win by 8/1 according to CNN & Fox
400 short that Arthur will be the strongest Hurricane of the season
400 short that Michael Strahan will play for someone other than the Giants
382 long that the Amazing Race 13 will premier in September
300 short that Wikipedia will hit 2.5MM English articles by 6/15
250 short in will McCain pick a VP today
250 short that Alexander Frei will score the most goals in Euro 2008
250 short that Rogers Clemens will be doing anything other than “something else” or “defending against perjury charges” on July 1
250 short that the US will have an Avian Flu outbreak before 7/1/2008
250 long that Hillary will speak at the Democratic convention
250 short of every stock in the Battlestar Galactica “Final Five” market except for the 4 already identified members

Topic: Feature Requests / Large Trades Distort Prediction Accuracy

Large Inkling Accounts distort prediction percent values, and thus accuracy, in a way that doesn’t reflect how physical markets (Currencies, Bonds, Grains, Energy, Stock, etc.) trade, or surveys would predict. When larger volume players enter a market by pushing a large size into a single transaction, the market will spike in the direction of the trade and stay there. This is no different than how the Inkling markets react, but what is different is that in the physical world markets, the prices revert to a value that is near to where it was prior to the clumsy way the position size was entered into the market. This is especially true since the electronic markets have now become the dominant way of trading in real world markets. If the Inkling markets were judged based upon a daily survey, the bias of the survey would not react to the large voice of a single person, but would count that opinion in comparison to how it favored the collective opinion’s bias.

In the case of how our Inkling markets react, the volume is interpreted as a larger more significant opinion of what will result, when in reality all that is being reflected is someone who is willing to bet a large size because they have a larger balance. There is no significant positive correlation between bet size and prediction accuracy, so trade-size shouldn’t be a controlling influence on prediction bias. Instead, a lot of the same opinions should be the dominant predictor and that only begins to have significance once a larger enough sample of opinions have been collected.

From watching how markets are playing out, larger size doesn’t improve the prediction, but it does push the prediction pricing around enough that it creates exit signals for prudent players, forcing them to leave the market because of the prediction’s price impact. It also distorts the market’s possible win/loss ratio enough to make the market unattractive to participation, or even to stay with the current position of the market. Taking new, or staying with current trades where the current Win/Loss ratio can cause a negative distortion to the trader’s positive expectancy percentage is a critical factor to avoid for a player/trader’s survival in the real world, and it is no different for a prediction player.

Instead of leaving the pricing so distorted and unrepresentative of accumulated opinions, it would be closer to reality that when a position size is applied to a market, that trade’s pricing should continue to be priced based upon how large the accumulated volume is impacting the current market’s prediction value, which is what happens now. However, and after the transaction is completed, the prediction price should then be allowed to return to a level near to where it was previously so that other participants in the market will understand the pricing spike was the result of a large volume single transaction of one player having an opinion, instead of a large shift in the opinion’s of the collective participants.

In simple terms, as more people accumulate volume on one side of a market, that should be the mechanism that moves the prediction in a direction, not the single size of a single bet size. To do otherwise changes the dynamics and reduces validity of the prediction markets. It also increases the risk of participating in a market because large volume trades are not viewed as aberrant price spikes.

I know that by pushing the prediction pricing around in Inkling’s current brutal fashion, people are avoiding participation in markets because the risk versus the reward of participating is too far out of line of what is acceptable. Low participation in the markets can only reduce the value of a prediction market, which doesn’t serve anyone’s best interest.

Please consider applying some average size weighting for adjusting the prediction pricing so that the collective opinions are more representative of how the market’s participants believe the question will be resolved.

Topic: Feature Requests / Performance Graph

I second a performance graph – it would be helpful.

Topic: Feature Requests / Standing Position Stop-Exit & Open-Risk

Great News Nate!
While I’m on a roll with this, I would like to add one more item to consider – Limit Orders for Exits/Entries.

When a position moves in a favorable direction there can be an advantage available to a trader to take the money and move on to something else, or close out the position and enter at a more favorable level. This is happening now in a couple of my current positions where the probability my initial direction guess was correct is increasing in probability. In this type of situation it seems like it would make sense/price-points to close out the position as it accumulates a predetermined price level and then consider entry at a more favorable calculation level for market termination.

Using Limit Orders for predictions would also expand the number of markets I would be interested in if there was an available Enter-at-Limit price option. In some cases, it is hard to see an advantage in taking a position when market sentiment is so bullish or bearish that the opportunity to a favorable conclusion is worth the risk at that point in time. In those cases, I’ve seen markets where I would take a position if the sentiment were at a more favorable level.

This means that if I could have placed a Limit-Order to enter either long or short (assuming assets available) at a user selected price Limit, this would allow a trader to enter a position if the probability moved to where a trader might see a better risk-reward ratio, without them having to grind through the long list of markets each day to update the current market’s probability level.

Have there been any thoughts on these kind of trading orders?

Topic: Feature Requests / Standing Position Stop-Exit & Open-Risk

Yes, it would be a great feature and thanks for the input you guys. It’s on a list of things to consider getting done soon.

Topic: Feature Requests / Standing Position Stop-Exit & Open-Risk

In principle it is possible to integrate these kinds of bids/offers into the kind of automatic market maker that Inkling uses, but my understanding is that it very significantly complicates the operation of the market. (Currently, the market maker can calculate directly the price effects of various offers. I think to integrate standing orders would require the market maker to combine the formulaic calculation with database look-ups and adjustments to the calculation for each order found, with a resulting significant slow-down in performance.)

However, I’m just tossing in my two bits; I don’t speak for Inkling.

I can see the value in what your propose and would be interested in hearing Inkling’s response.

Topic: Feature Requests / Standing Position Stop-Exit & Open-Risk

Not being able to have an active standing order to Close-Out a position at a trader selected price is an uncomfortable way to trade, even though we are using virtual money.

Not having a Position’s Risk-Value, and an Account Open-Risk value keeps the presence of risk in the background.

If a standing exit order along with the risk to that order was available, it would allow position trading to be more in line with how professionals trade real markets with real money. This would then allow position sizing decisions to be based upon initial position risk and then allow the trader to move the standing Close-Out order as the market moved in their direction, or be out of the market whether they were able to access the site or not.

Are these options being considered at some point?

Topic: Feature Requests / Dividents

The suggestion for adding dividends for long term markets would have the effect of creating an investment incentive in the long term markets to compensate for Inkflation. Instead it seems the focus should be on stopping Inkflation.

Topic: Feature Requests / Performance Graph

Ahh, yeah, that makes sense. Thanks for the recommendation.

Topic: Feature Requests / Performance Graph

Yes. Please. That would help so much. I have the hardest time remembering what my specific stats were during my last visit.

Topic: Feature Requests / "Most recently traded" in the markets list

Right, agreed. It’s on the task list.

Topic: Feature Requests / "Most recently traded" in the markets list

I was searching for markets the other day and come across some that hadn’t seen activity in a long time even though they’re still quite pertinent to current events. Without doing a search by keyword, I never would have seen them. Perhaps a “most recently traded” or a “traded today” listing might make it easier to revive some of those auctions that involve an event in the semi-distant future.

Topic: Feature Requests / Dividents

My thought on the moderator really was more focused on trying to ensure that the market was set up such that trading was suspended before the results were known. In other words, for individual games, ensure the market is set up to automatically close before the known start of the game. The formula 1 market, for example, was set up that way – trading closed shortly before the event started. Some markets with multiple outcomes would be more difficult to do that with. Still trading on college football, for example, could be suspended every Saturday until results are known. Having not run a market, I don’t know if that is even possible…but it is a thought. Making it easier for the moderator to be able to figure those things out would be helpful as well.

With respect to the margin comment, my thought in your example would be rather than having to reserve 350 inkles, you could meet that requirement by having maybe at least 700 inkles in “value” of other stocks in your account allowing you to put that 350 to work elsewhere. Set it up with automatic margin calls/liquidation of positions should the excess value fall which is the risk shorts or any margin player takes anyway.

In my Argentina example, the stock should be at 0 but is ~8 at the moment. However, to get at those 8 inkles, one has to completely tie up 92 which will probably happen the day the stock closes (or if a millionaire choses to impose discipline) but, for the reasons noted above, probably not before then. I might be able to personally drive it to 0 if the “value” of my other investments were considered but I’m not interested in tying up more inkles until the market closes on Monday when I can make inkles in the meantime in other markets.

You may be right about the wisdom of handing out more at the start but without enough liquidity it is tough to stimulate more active trading. I do like your limits by the way, but unless there are more traders and/or activity, the markets will tend to remain inefficient (though the inefficiency in itself may encourage more trading with just some of your suggested limits).

Topic: Feature Requests / Dividents

Even though my thoughts were written with being a marketplace manager of my own in mind, I agree on the problems you arise concerning the inklingmarkets market place.

Indeed at inklingmarkets, closing a market is a huge problem but the moderator can’t possibly know all events happening around the world.
Therefore I suggest to force a market owner to add a URL of reference when creating a market. As such the moderator can go to that URL and verify closing and possible stocks.
Next I would suggest to add an extra page to the wizard to create markets that is specially dedicated to closing the market. Asking the market owner several dates for closing the market upfront will have a smoother experience for traders. Several dates are required for markets like tournaments where stocks are closed at $0 when a team loses a game.
I also suggested somewhere here on the forums to add timezones to the system so that market owners do not have to calculate their time agains the time of the site.

Payout is another problem that needs to be dealt with. If a trader has won, he doesn’t like to wait for weeks to be payed out…

We have to make the trading experience as good as possible so that more traders stick around and keep trading…

The initial amount of inklings is indeed a problem… After a couple of minutes all my money was invested on only a few stocks. I even didn’t know yet that I have to carefully look at the closing date so my money isn’t blocked for a too long time… As such I had no more money to spend and I had to wait a long time for results… This will force a trader to leave the marketplace and “forget” to come back…
The first 2 of my suggested investment limits could have solved my problem as I would have been investing less money on a single stock and market, leaving me money to invest some more.
I do not believe that giving more money is the answer because that will result in some more investing whithout thinking and experience when I start.

The “going short” solution you give won’t change a thing. If you go short on a stock that is worth $30 and you sell 5. $150 will be deposited in cash and $500 reserved which is the same as reserving $350 like you suggest…

Topic: Feature Requests / Dividents

The basic problem is a combination of 1) there are not enough traders and not enough depth in the markets to make them truly efficient and 2) given how some of the markets are run (e.g. not closed until the result is known) there is too much “free money” available in the short run to make participating in longer markets attractive almost irrespective of the “interest rate” one might pay.

I’ve not been though most of the markets but there is at least one I am participating in that is essentially a lock to more than double in price within the next 6 months however the payout is at least that far away. With the ability to more than double ones money over a weekend by doing football or baseball games in the 4th quarter, 9th inning or even after the game, given limited inkles to invest, what is the point of investing in a longer market? Coupled with the large tie up of inkles if you want to short makes it even harder to play long term markets if you think the positve answer is wrong.

Heck, there is a market asking whether the current President of Argentina will be re-elected. The guy isn’t even a candidate, and hasn’t been one for months. I suppose he could encourage a massive write-in campaign that might turn the tide, but his wife, who is the leading candidate and his presumptive political heir, would probably be pissed. An efficient market would have pounded that baby down to ZERO months ago. Admittedly the question was posed in Spanish but anyone that found this site can probably find babelfish and any search on the name would tell you the answer to the question posed. I’d drive it to close to zero (I’ve contributed to its decline) if I had unlimited resources but I don’t have the inkles to do it. I’d rather go with the quick pops.

Which leads to a couple of suggestions:

- go with investment limits suggested by ednique which would at least stop the concentration of resources in short term “sure things”
- have the moderator ensure that markets are set up to halt trading before sporting events begin, before election day, when the stock price is within 5% of target – in general before the outcome is reasonably certain. (Some market makers do a nice job of this in market design – some don’t close out markets until days ater the result is known).
- combine the above with giving people more inkles to begin with to stimulate more active trading.
- change the short rules to some kind of margin requirement rather than “inkle reserve” difference between 100 and the price.

All of these things might encourage more participation in longer term markets, while the last three should tend to make the markets actually more “predictive” rather than a game of who has the fastest internet connection to sell 50 shares of “LSU goes undefeated” in the final seconds of the game.

Topic: Feature Requests / Performance Graph

A performance graph in the dashboard would be great.
As a matter of fact, it may even appear on the public profile.

Topic: Feature Requests / Dividents

Very good remark onemike…
If P1 closes this week and cashes out $100 and P2 pays a dividend of $5 this week and closes next week cashing out $100, I think a trader would buy P1 this week and a similar as P1 next week…

Concerning the predictive value of the market I fear dividends will influence the market in a bad way.
If a stock is at $25 and a dividend of $5 is paid, the stock goes down $5 to $20. If there are other stocks in that same market that do not get a dividend, their value will increase. As such the market gets flattened out after dividend.
Offcourse, people can trade again, but the results as before the dividend isn’t there anymore…

• set limits on money invested in a single stock as a % of their assets.
• set limits on money invested in a single market as a % of their assets.
• set limits on money invested in 1-week-markets as a % of their assets.
• set limits on money invested in 2-week-markets as a % of their assets.
• set limits on money invested in 3-week-markets as a % of their assets.
  They could still avoid long term contracts and keep more savings…

Topic: Feature Requests / Dividents

Dividends or ‘interest payments’ of some sort would encourage participation in longer term markets.

I wonder what discount rate or inflation rate would be appropriate if interest were to be paid on account balances. If some sort of return was paid for holding a long position in a contract, would a short position similarly be paid a return or would the short position be charged?

I’m not sure about the economic issues and how they work out.

Some basic analysis might go: assume a trader has a balanced position in two contracts (balanced meaning trader’s prediction = existing market price for both contracts P1 and P2) one which expires in one week and the other which expires in two weeks, next, assume new information about the later contract such that the trader now expects the price to become P2+X. How big does X have to be to induce a trader to sell contract 1 to fund additional purchases of contract two? I.e. how much return next week will a trader give up to get a larger return in two weeks?

In principle it seems possible to set up such pairs of markets and then infer the amount X from trader behavior.

Knowing this amount doesn’t answer the question of whether it would be helpful to pay dividends or interest on longer term markets, but it would be one thing you would want to know.

Topic: Feature Requests / Dividents

It seems that short term markets have very active trading but the long term markets (almost) nothing…
As such I think that paying out dividents could help increase trading in those markets.

If one would receive for ex $5 each week, he will probably trade in a market that lasts a year. If not, his money will be blocked during a whole year and he won’t trade in that market…

How this would work, I don’t know, but I thought that you could allow the market owner to pay divident to a single or all stocks in a market.

If you have a market about a championship, you could award each team that wins a game $5 divident. Or if you are guessing a date, $5 each week the event has not been taking place.

As in normal stock markets, when paying dividents, the stock price is lowered accordingly.
Once again other traders could as such see new opportunities in a stock after devident.

As I’m not a specialist on the subject I have no clue what paying dividents mean in prediction markets in terms of their predictive corectness…

Topic: Feature Requests / abbreviation and trades on market view

It would be nice to see the abreviation of each individual stock in the market view… This could make the graph more clearer

All lists display current value, todays result and number of trades. Once in the market view the number of trades are not shown anymore…