Decentralized prediction markets may be the answer to unreliable information, which is a pernicious issue that affects not only our daily lives but also the way markets behave. We may learn that a stock is going to drop by an expert or a media source we rely upon. But then it turns out to be false information because the “facts” are derived from only a small set of people.
There are a few serious issues that contribute to the problem of misinformation. One problem is that we are forced to rely on authorities for quality information: our financial advisors, our friends, the news. And, all of these aforementioned individuals and institutions have certain limitations making it impossible for them to have all of facts.
To make the problem worse, when we act on bad investment advise and misinterpret the market, it costs the investor. It does not affect the website, person or institution from which we received the information.
This issue is not new, and it is a problem that decentralized prediction markets aim to solve.
Wisdom of Crowds: Prediction Markets
In response to the obvious limitations of siloed knowledge, the “wisdom of crowds” or “crowdsourcing” has increased in popularity. Predictive markets are proving that truth and facts have economic value.
Prediction markets have been around for a while, and many economists see them as a very valuable source of insight into market behavior. Presently, decentralized prediction markets are primarily used to predict the outcomes of sporting and political events. But as decentralized prediction markets continue to grow in popularity, they will increasingly deal with more types of predictions, such as the case with HedgeTrade, where trade predictions are made on multiple cryptocurrencies.
The main reason prediction markets and the wisdom of crowds are gaining popularity is because there is more information gathered from more people.
How predictive markets work
So here is how prediction markets work; information is aggregated from many participants. Based on what this information is for, maybe a flu outbreak, or a weather phenomenon that will affect crop production, the market generates a price. The price is simply a response to what this information is worth. Prediction markets not only compile better predictions than individuals, but people can then trade the outcomes of these events.
In the last 15 years, there has been an increase in economists interested in aggregated information. The data is collected from multiple groups of people.
Evidence suggests that groups tend to outperform individuals when it comes to aggregating information, weighing alternatives, and making decisions.
This means that groups are better at predicting the outcome of an event. The reason for this is that there are more participants all sharing their individual expertise or insights. This is then aggregated to reflect the most likely outcome of an event. Therefore, many economic theorists believe that prediction markets are highly beneficial for predicting market behavior.
Skin in the game
One good reason for the improved accuracy of prediction markets is because in order for an investor to participate, they must stake a bet. And because there is money on the line, this has been shown to incentivize better research, good behavior and intelligent decision making. If you are correct about the outcome, you will be rewarded. If you are wrong, you will lose your stake.
As an additional bonus, because there is a potential reward, prediction markets have increased participation. That means that more information from more subgroups contributes to a more holistic interpretation of information. So, because there is a large rate of participation, the accuracy of the information sourced is higher than when data from smaller sample groups is used.
Imagine how much more accurate the predicted success of a piece of technology would be if everyone involved in its design and rollout were able to anonymously participate in its success (or failure for that matter).
And because it is anonymous, participants can be truthful. Since it is a secondary market, even if the product fails, it has created a secondary market which profited off the predicted failure.
What is a Decentralized Prediction Market?
Decentralized prediction markets are poised to provide even more accurate information about world events and markets. This is because they source information for predictions from a borderless population. With more participants and multiple sources, the accuracy of the information provided should improve appreciably.
With the inception of blockchain and cryptocurrencies, decentralized prediction markets (DPMs) have become a reality. We’ve seen a number of DPMs building off the Ethereum platform. One example would be Augur, where you can participate in predictions for political outcomes, sporting events, and who Warren Buffet will take to lunch.
Another example of a decentralized predictions market is HedgeTrade, where traders earn by publishing Prediction Blueprints for cryptocurrency trades.
With HedgeTrade, participants use a tokenized app to make market predictions on various cryptocurrencies at different exchanges. The main idea is to enable professional traders to sell their predictions while also staking differing amounts as a way to earn even more.
But what sets HedgeTrade apart from its competitors is that it holds its traders accountable for their predictions in two ways. If a prediction is incorrect, they lose their stake. Additionally, those that bought their prediction get their money back.
So the HedgeTrade app helps to solve the problem of accountability in information disbursing. On top of that, you still have the aggregated crypto trade data plus the borderless and anonymous features that make prediction markets so attractive.
It’s very likely that platforms such as these will have a very positive effect on the increased success of market predictions. And as the aggregation of collective information continues to improve, it will also have a positive effect on the struggle between truth and fake news.
The new era of DPMs
With the increased popularity of cryptocurrency, decentralized prediction markets have entered a new era. In their current form, however, as with cryptocurrency and blockchain applications, there are some technological limitations, such as with scalability and application. But, just like cryptocurrencies, which have exploded in the last 10 years, we have good reason to believe that DPMs will also increase in popularity.
What Incentivizes Participation?
Improved quality of public information is not the only reason to participate in prediction markets. To participate in a DPM one must stake money; you must literally put your money where your mouth is. So, if your prediction is correct, you are rewarded with a payout. If it is not, you will lose your stake.
So participating in a prediction market means buying shares in an event outcome. The kind of future events to place a stake on are enumerable. But they are limited to yes or no outcomes, or correct or incorrect. These predictions are known as “atomic” because they are binary (yes or no) or they occur or do not occur.
- In a prediction market, price equals perceived probability. That means that if a YES share costs 75 cents, this means that the market (and collective) believes there is about a 75% chance that the event will occur.
- And if each NO share costs 40 cents, then the market thinks there is about a 40% chance that the event will not occur.
The future event could be if it will rain in Paris tomorrow. Or if a stock’s value will increase, or who will win a forthcoming election. And because there is a large amount of participation in the outcome of real-world events, more information is aggregated. That means that more factors are considered when trying to determine an outcome.
So if another prediction is that Apple’s stock will increase, and then the developers share that there is a problem with a new product, they can actually bet against the stock going up. As a result, not only does the rest of the market respond faster, but a secondary market of predictions is created. So if Apple’s stock drops, but many employees predicted this correctly, then they will benefit financially from being correct. Even if the company does not benefit from it.
Finally, when you participate in a protocol such as HedgeTrade or Augur, you not only stake your money on your prediction, you also stake your reputation. That means that if you consistently offer poor, uninformed advice, your reputation is proportionally affected by this bad judgment. As a result, your word will become less valuable, having less meaning for the results of any future prediction.
In a certain way, prediction markets are creating a new market out of uncertainty, while simultaneously improving the share of information.
How does this all work?
Decentralized prediction markets work because there are many participants all around the world. This broad-reaching and diverse collaboration inevitably contributes to a stable network. It is also established on a blockchain, so its immutable cryptographic ledger ensures that information is maintained accurately and securely.
The reason this works is that in order to participate in a decentralized prediction market, you need to hold cryptocurrency. Therefore, if you want to participate in a prediction you need to stake crypto on it.
Financial reward and risk of failure demonstrate a marked improvement in the reliability of the information.
When people stand to risk money they are more likely to better prepare themselves by doing more research in an effort to increase their chances of success.
Moreover, information is sourced from a much larger pool of participants.
With surveys and census data collection, the information inevitably comes from a small selection of people – experts in the field, and easily accessible willing participants. In both cases, this means that the behavior of markets and world events is limited to a small subsection of people based on social demographics, geographical location, as well as the ease of accessing these people.
Because of this, those who are surveyed will often cite the same information. However, if the section is broader and globally based, then those participating are going to bring a broader demographic of information to the table. Based on the collective of information, this would allow for more accurate predictions.
Limitations of Centralized Prediction Markets
Prediction markets are not really a new thing, but until now they have been ‘centralized.’ What we mean by centralized is that predictions are based on information that is gathered by a specific group. It could be a team of research scientists, statisticians, actuaries, political groups, private companies or personal interest groups.
Although none of these groups are intrinsically bad for markets, the problem is that they all have their own limitations. The primary problem is that each group has a specific interest that leads to a certain limitation of their collected information. And then only like-minded people will seek out the information that is provided.
Both private companies and political groups are going to be interested in outcomes that serve their economic and power interests. If the information is collected by academics, the results often end up siloed in universities and published in cryptic language, inaccessible to non-academics.
Thus the overall problem of centralized information is that it is limited to those with privileged access, as well as those lucky enough to have their research interests funded and supported. As a result, the accessibility of quality information continues to be a challenge.
How data sharing is better with DPMs
Centralized prediction markets in some shape or form have been around for a long time. What decentralized prediction markets bring to the table is improved access to information sharing.
A prediction market can be compared to Wikipedia. Just as anyone can add something to Wikipedia’s free encyclopedia, so can anyone with more or better information report a problem or add to an entry. The information from the encyclopedia is globally shared, the only limitation is access to a network.
The idea is not only are two heads better than one, but that there is a shared responsibility to ensure accuracy.
Beyond the improved sharing and accuracy, DPM’s add in financial incentivization. So while you might be tempted to have a laugh on Wikipedia and say that the capital of Canada is New York City, you would not want to fool around with the data you are entering on a platform like HedgeTrade or Augur.
Because on the DPM, you must stake your money and your reputation. A smart contract is in place to make sure you only get paid for accurate information. And you will not gain any agreement that NYC is the capital of Canada (because it is Ottawa, by the way). As such, your reputation score would decrease relative to the bad information you submitted.
Here’s how it works:
As mentioned, Buy and Sell transactions are atomic, which means that there are only two possible outcomes of the prediction. The predictions are settled with smart-contracts. Therefore, as soon as the conditions of the contract are met, the transactions (which are payments) are executed.
Both decentralized prediction markets and Wikipedia demonstrate the “wisdom of the crowd” phenomenon. The main difference between the two is that in a prediction market, accuracy is capital. So you have nothing to gain and everything to lose from perpetuating false information.
And, not only do DPMs create liquidity out of accurate information, before DPM and Wikipedia, access to quality information was also limited to those with access to private academic journals, physical libraries, and expensive expert advice.
DPMs are a step towards fewer barriers to quality information, as well as fewer limitations on sharing information.
Challenges and Limitations
While there are many potential benefits to DPMs, there are also some limitations that these projects face. Ethereum and other blockchain technologies may be limited to the extent so they don’t have a sufficient level of participation. Decentralized markets are still very young.
That said, Wikipedia started off in some pretty dire conditions and has since become a useful and accurate source of information and responsible peer review. So, it is entirely likely that in 5 to 10 years the prediction markets will be much more far reaching and influential, as well as profitable, to its participants.
On that note, like all blockchain protocols, DPM faces scalability issues. Using blockchain means relying on an immutable cryptographic ledger. However, speed and scalability are very limited to maintain blockchain’s security. Developments on multiple blockchains are well underway to try and solve this problem.
Moreover, as of now, prediction markets face regulatory limitations. They are considered a form of gambling in several American States. And as cryptocurrency and decentralized networks gain popularity, they also face newly mandated regulatory bodies. Facebook’s Libra is a prime example of the institutional response to an big tech-sized cryptocurrency project.
Some economists have suggested that in an effort to manage the worry of gambling that there should be an individual cap on what one can spend in the predictive market annually. No doubt, however, this rubs against many with libertarian ideals that are drawn to cryptocurrency.
Fear of collusion
Finally, and the genuinely concerning limitation of any prediction market, is the fear of collusion. This is when there is an organized program which intends to deceive and mislead the markets. It is not hard to imagine that companies or individuals with enough means would want to use their money to influence decisions about stocks.
Incentivization and Problem Solving
Incentivization is at the core of the solution to many of the challenges that face information dissemination and prediction markets. Those who stake on the accurate outcome win an additional stake. Those who report inaccurate outcomes will lose their stakes.
Many economic studies demonstrate that these stakes are an effective way to incentivize honest reporting and personal fact-checking.
The overall design of decentralized systems harnesses the incentivization of honesty with monetary rewards. Decentralized networks like the Bitcoin Network only work when there is a 51% agreement. Similarly, prediction markets only work with a combination of blockchain, cryptography, and real-world events; if the event doesn’t happen then it is not recorded in the blockchain.
One of the reasons for the success of some decentralized prediction markets is that they are simple. There are only two kinds of predictions: YES (long) shares and NO (short) shares. The payout of each share depends on the outcome of the event, and so payout takes place only after the event has occurred and an outcome has been reported.
In a simple prediction market, each YES share pays out a dollar if the event in question occurs. Alternatively, there is no payout if it does not occur. But, if there is no reported result, then a NO share pays out a dollar.
With HedgeTrade, you have certain choices when making a prediction:
- Which cryptocurrency
- What exchange
- The amount to stake
- How much to charge for your Prediction Blueprint
- Your expiration date for the prediction
Predictions on the HedgeTrade Platform are related to certain cryptocurrency prices as traded on various exchanges.
What you need to participate
In order to use a DPM, you will need a digital wallet with cryptocurrency, bitcoin and Ethereum are great to start you out. You can use any of the following industry standard wallets of your choosing, such as MetaMask, Ledger, Trezor, AirBitz, and uPort. A wallet is necessary to that users can participate in transactions and coin staking.
Review of DPMs
Prediction markets are just as they sound; a market made out of predictions formed by collective knowledge. Essentially, prediction markets monetize information. However, only accurate predictions are rewarded, while inaccuracy will cost you.
DPM protocols are currently built on Ethereum include HedgeTrade, Stox, Gnosis, and Augur. Related protocols on other blockchains include Hivemind built on Bitcoin, and Bodhi built on Qtum.
Decentralized Prediction Markets offer a promising new market.
Here are a few of those reasons:
- DPMs are not limited by location or citizenship. That means that anyone with network access and knowledge to share can participate. This is because decentralized predictions work with blockchain protocols that are open and public.
- DPMs use a model of incentivization by monetizing an event. So if you want to participate in the market prediction, you need to put your money where your mouth is. Evidence from multiple economists suggests that staking money incentivizes knowledgeable contributions and de-incentivizes uninformed opinions.
- DPMs are possible because of the application of blockchain and its use of cryptographic blocks and cryptocurrency. This means that you must hold cryptocurrency to participate in predictions. Moreover, all results are settled using smart-contracts, which means that they are executed automatically and are based on the outcome of the prediction.
- All Buy and Sell transactions are atomic, which means that there are only two possible outcomes of the prediction; the even occurs as predicted or it does not.
- DPMs are third-partyless and without regulatory bodies. The only regulations are built into the program and the smart-contracts. Also, in many states, centralized prediction markets are considered a form of gambling and are therefore illegal.
- Value and prices of shares are determined by the market alone. Decentralized means deregulated, so the value of a prediction is based on the need and perception of the market. The equilibrium of the market is represented by the present perception of a market and the information gained from the crowd.
DPMs have a great deal of potential for harnessing and improving the wisdom of the crowd. Information is harnessed without borders, from multiple locations. As a result, the information is gathered from a much larger group of people.
So, not only do DPMs offer a broader range of information, but they also reward better information. By offering a reward, not only is honesty incentivized, so is improved research. Most importantly, dishonesty comes with a financial loss. This means that there is potential for a renewed market on research. Performing research studies is expensive, time-consuming and only ever offers limited results based on the control groups studied.
But if good predictions are valued like assets, then everyone has a vested interest in forming better predictions for the success of consumer products, the effectiveness of prescription drugs, and the availability and scarcity of food, to name a few possibilities.