Paper: arXiv 2310.05971
Authors: Victor Olkhov
Abstract
This paper studies the links between the descriptions of macroeconomic variables and statistical moments of market trade, price, and return. The randomness of market trade values and volumes during the averaging interval Δ results in the random properties of price and return. We describe how averages and volatilities of price and return depend on the averages, volatilities, and correlations of market trade values and volumes. The averages, volatilities, and correlations of market trade, price, and return can behave randomly during the long interval Δ2»Δ. To describe their statistical properties during the long interval Δ2, we introduce the secondary averaging procedure of trade, price, and return. We explain why, in the coming years, predictions of market-based probabilities of price and return will be limited by Gaussian distributions. We discuss the roots of the internal weakness of the commonly used hedging tool, Value-at-Risk, that cannot be solved and remains the source of additional risks and losses. One should consider theoretical economics as a set of successive approximations, each of which describes the next array of the n-th statistical moments of market trades, price, return, and macroeconomic variables, which are repeatedly averaged during the sequence of increasing time intervals.
Complexity vs Empirical Score
- Math Complexity: 6.5/10
- Empirical Rigor: 2.5/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper relies on heavy mathematical notation involving statistical moments and successive approximations, but it lacks backtest-ready implementation details, specific datasets, or empirical validation, focusing instead on theoretical frameworks.
Research Flowchart
flowchart TD A["Research Goal:<br>Link Macro Variables to<br>Microstructure Statistical Moments"] --> B B["Methodology:Sequential Averaging & Approximation"] --> C C["Input:Market Trades<br>(Values, Volumes)"] --> D D["Process 1:<br>Averaging over interval Δ"] --> E E["Output 1:<br>Random Price & Return<br>with moments dependent on trade moments"] --> F F["Process 2:<br>Secondary Averaging over interval Δ2 >> Δ"] --> G G["Key Findings & Outcomes:<br>1. Limitation of Gaussian predictions<br>2. Internal weakness of VaR<br>3. Theory as successive approximations of n-th moments"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a market‑based approach to price and return probabilities that incorporates the randomness of trade values and volumes during the averaging interval Δ and links statistical moments of price/return to moments and correlations of market trade.
Source sentence
The market-based approach to price probability that we propose takes into account the randomness of trade values and volumes during Δ and establishes the direct dependence of price and return statistical moments on statistical moments and correlations on market trade values and volumes.
Assumes 6
- Trade frequency is very high, with intervals ε as short as a second or less. market
Source sentence
However, the frequency of trades in stock markets, FOREX, commodities markets, etc. is very high.
- A time‑averaging interval Δ is chosen such that ε << Δ << T for long‑horizon analysis. model
Source sentence
To do that, one should choose the time interval Δ, such as ε<< Δ <T, and aggregate or average the market trade time series during Δ.
- Trade values C(t) and volumes U(t) are random during Δ. statistical
Source sentence
The market-based approach to price probability that we propose takes into account the randomness of trade values and volumes during Δ and establishes the direct dependence of price and return statistical moments on statistical moments and correlations on market trade values and volumes.
- The initial high‑frequency trade series are irregular and need to be aggregated over Δ to obtain smoother dynamics. model
Source sentence
The choice of the interval Δ and the aggregation of market trade time series during Δ smooths the irregularity of the initial trade time series and determines the collective impact of market trade on agents’ and macroeconomic variables.
Show 2 more
- Prices are given by the identity C(t) = p(t)·U(t). model
Source sentence
Each market deal with the trade value C(tik) and volume U(tik) at time tik defines the price p(tik) due to a trivial equation: 𝐶(𝑡𝑖𝑘) = 𝑝(𝑡𝑖𝑘)𝑈(𝑡𝑖𝑘) (3.4) For convenience, we consider all prices adjusted to the current value at time t0.
- The first statistical moment (VWAP) is used as the market‑based average price a(t;1). model
Source sentence
As the market-based average price or the 1st statistical moment we take VWAP (4.2).
Data
No dataset named in the text.
Costs
Not applicable
Finds 5
- Predictions of price and return probabilities will remain limited to Gaussian‑type distributions for many years. theoretical
Source sentence
Thus, Gaussian-type probabilities of price and return are the only possible predictions for many years to come.
- The market‑based approach links price/return probabilities to the stochasticity of market trade, highlighting the origin of successive approximations in economic theory. theoretical
Source sentence
That links market-based price probability with market randomness and macroeconomic evolution.
- Value‑at‑Risk (VaR) relies on frequency‑based probability assessments that ignore random trade‑volume effects, creating additional risks and losses. negative
Source sentence
Frequency-based assessments of price or return probabilities ignore the dependence on the random size of trade volumes. The differences between the market-based and the frequency-based approaches to price and return probabilities are the source of the additional risks and extra losses.
- Even market‑based probabilities are constrained by Gaussian approximations due to current limits of economic theory. theoretical
Source sentence
As we discussed above, the current state of economic theory at best limits the accuracy of any predictions of probabilities of price and return by Gaussian distributions.
- The paper shows that higher‑order statistical moments (beyond the first two) require development of higher‑degree economic models, which are currently unavailable. theoretical
Source sentence
The development of 2nd degree economic models may take years, but it is the only way to improve the predictability and reliability of economic theory. Predictions of the higher statistical moments require the development of economic models of the 3rd degree, 4th degree, etc.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 23 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.