The Trump Chaos Index (TCI) is a real-time gauge of the severity of policy uncertainty associated with the second Trump administration.
TCI is derived from dozens of prediction market prices on legislative outcomes, staffing decisions, executive actions, and more exotic things like impeachment and removal from office.
A detailed methodological discussion is available below, but suffice it to say a higher number suggests more chaos, less stability and less visibility into short/medium-term policy changes.
More significant than TCI’s absolute level are its direction and rate of change over time.
Discussion
Each component in the TCI is classified as chaotic, neutral, or antichaotic.
A chaotic component implies more policy uncertainty as its odds of occurrence approach 100%.
A neutral component implies more uncertainty as odds approach 50%.
An antichaotic component implies more uncertainty as odds approach 0%.
Classification does suggest a normative evaluation of an outcome as socially desirable or undesirable, economically beneficial or harmful, or otherwise good or bad.
A chaotic component is simply one that, if its underlying event were to occur, would represent significant, perhaps even unprecedented political upheaval, such that it would necessarily usher in multiple significant follow-on policy uncertainties (e.g. eliminating a federal department or repealing landmark legislation).
A neutral component is one tied to an outcome that different groups might favor or disfavor, but whose outcome - if relatively assured, whether affirmatively or negatively - is well precedented, well understood, and/or can be readily quantified in terms of its economic or political impact (e.g. tax cuts, narrow regulatory reforms). Hence uncertainty is minimized when the odds approach either 100% or 0%.
An antichaotic component may simply be the mirror image of what would be a chaotic component (if the prediction market question had been written in the negative) or it may be tied to a multi-outcome market that “should” resolve to a telegraphed preferred outcome, such as an announced nominee being the candidate ultimately confirmed to that office. In the case of multi-outcome markets, the highest-priced outcome is used in constructing the index.
Each underlying outcome is assigned a subjective relative weight (detailed below) to reflect the fact that not all chaos is equally utter. Because prices are drawn from multiple real-money prediction markets, when two or more markets are trading the same or a sufficiently similar outcome, that outcome’s weight is split across the multiple equivalent markets.
Construction Methodology
Prices are collected once per minute and represent the midpoint of the bid/ask spread published at that time by the respective platforms.
TCI = 100 * sum of Weighted Chaos Scores (WCS) for each market.
WCS = Unweighted Chaos Score (UCS) * Weight / sum of Weights
UCS = Price (for Chaotic components), 1 - Price (Antichaotic), or -2 * |Price - 0.5| + 1 (Neutral).
We rebalance the index from time to time and for various reasons, including the launch of new markets and resolution of existing ones. Whenever we rebalance, we apply a normalization factor sufficient to cause the immediately pre-rebalancing and immediately post-rebalancing index levels to equal one another. That normalization factor is applied to each calculation until the subsequent rebalancing, when the normalization factor is recalculated.
This is awesome. We need more Chaos Indecis for other countries. Maybe I will make some one day
If only the old Trump Tweet markets were still around to help quantify the chaos.