Investing
Our approach to portfolio construction, analysis, and allocation.
This page describes the specific analytical framework we use when making investment decisions on behalf of clients. It sits inside our broader framework, which starts with the person rather than the assets. Here, we focus on the assets themselves — how we think about them, evaluate them, and combine them into portfolios.
Where the analytical work is implemented in our open-source project nexus-core, we reference the specific tools. Academic sources for each component are cited at pwos.app/references.
Overview is the plain-English version. Detailed has the full method — five regimes, seven durability layers, eight quality checks, with academic citations.
The method, plainly
Three questions, asked in the same order every time
When clients ask us how we make investment decisions, the honest answer is that we ask three questions, in roughly the same order, every time. Same client, different decade — same three questions. The questions are not novel. The discipline is in asking them every time, in the same order, and writing the answers down.
What environment are we in?
Markets move through distinct phases. A given asset that does well in one phase often does badly in another, and the phase changes for measurable reasons — interest rates, credit spreads, volatility, currency strength. Before we evaluate any specific asset, we classify which phase the market is in. This is regime detection — classification of observable conditions, not prediction of where conditions will go next.
How durable is this asset?
Some assets compete on advantages that decay quickly — fashion, feature parity, the latest software release. Other assets compete on advantages that decay slowly — a regulated infrastructure position, a hard physical asset, a network effect that has been building for decades. We sort assets by how long their competitive advantage tends to last under stress, because the durability of the underlying advantage shapes how long we can hold the asset and how confident we can be about the holding.
Is this asset actually good of its kind?
Inside any category, some assets pass the basic financial-quality checks and some do not. We score each asset against eight checks — financial health, cash generation, trend persistence, technology cycle position, decay rate of the competitive advantage, fit to the current regime, sector composition, and a screen for AI substitution risk. Each check is anchored to an established academic source. Score and trend over time determine position size.
The same three questions are applied the same way to a stock, a bond, an ETF, a real-asset position, and a digital asset. Apple passes through the same scoring as Bitcoin. A Treasury bond passes through the same scoring as a high-yield credit fund. The point of the framework is that it does not change which asset class is in front of it.
Boundaries
What we don't do
The questions above describe what the framework is. It is just as important to describe what it is not, because every working method has a shadow — a set of things people assume it does that it doesn't. Five worth naming.
We do not predict markets.
Regime classification is a description of where conditions are today, anchored in measurable signals. It is not a forecast of where conditions will be next quarter.
We do not use technical chart patterns as primary signals.
Trend information shows up in the framework as one input among eight. It is not a buy or sell trigger on its own.
We do not treat Monte Carlo simulations as forecasts.
A Monte Carlo simulation describes a distribution of possible outcomes given a set of assumptions. It is a discipline tool, not a crystal ball.
We do not stock-pick as a primary source of return.
Most of the expected return comes from being in the right kinds of assets in the right environments. Specific name selection is a smaller, secondary contribution.
We do not give tax or legal advice.
Tax-aware portfolio construction is part of what we do. Specific tax recommendations are not. Those come from the client's CPA or tax counsel, working alongside us.
Foundation
Where this comes from
None of the underlying ideas are ours. Regime classification draws on Hamilton's regime-switching econometrics. The durability concept draws on Carlota Perez's work on technological revolutions. The eight quality checks draw on Piotroski, Mandelbrot, and several decades of factor-investing research. What we built is the integration layer — the system that takes those research-backed inputs, applies them the same way to every asset, and produces an answer the same way every time.
The integration layer is implemented in our open-source project nexus-core. Academic citations for each component live at pwos.app/references. You can verify what we say by reading what we wrote.
The Method
Three questions we answer before making an investment decision
The three questions are sequential. The regime tells us what kind of asset should be overweighted or underweighted. The durability layer tells us where this asset sits on the spectrum from secure to speculative. The quality score tells us whether to own it versus owning a cheaper alternative or no position at all.
Question 1
What environment are we in?
Markets go through regimes. Periods when growth stocks compound, periods when hard assets protect purchasing power, periods when debt-servicing capacity determines everything. Historical regime changes produce different winners, different losers, and different optimal allocations.
Growth
Risk assets rewarded. Liquidity expanding. Low inflation, rising productivity, strong dollar. Equities and innovation layers outperform.
Transition
Mixed signals. Elevated uncertainty. The regime is shifting — conditions that worked last quarter may not work next quarter. Selectivity matters.
Hard Asset
Inflation favoring real assets and commodities. Currency debasement. Physical infrastructure and energy layers outperform financial assets.
Deflation
Credit contraction. Falling prices. Liquidity crisis. Capital preservation dominates — cash and quality bonds outperform everything else.
Our regime-detection framework, the Entropic Macro Framework (EMF), classifies the current environment into one of five regimes: Growth, Transition, Hard Asset, Deflation, or Repression.
The primary signal is the ratio of gold to the S&P 500. When gold outperforms equities over trailing periods, we're in a Hard Asset regime and allocation should tilt toward assets that benefit from monetary debasement. When equities outperform and the ratio compresses, we're in Growth. Transitions between regimes are marked by specific signal combinations — yield curve behavior, VIX regime, real rates, credit spreads.
We do not claim to predict regime changes. We claim to identify which regime we are currently in, and to allocate consistent with that regime rather than against it. Regime identification is rule-based and repeatable; prediction is not.
The EMF framework is implemented in nexus-core under pw_get_regime
and related tools. It draws on work by Carlota Perez on technological revolutions and financial
capital cycles, and on a long line of research on regime detection in macroeconomic time series.
Detection Signals
Question 2
How durable is this asset?
We think of assets as having a characteristic decay rate. Good portfolio construction matches the decay rate of assets to the time horizon of the goals they serve.
Some things hold value for centuries (gold, productive land, ownership of irreplaceable businesses). Some things hold value for decades (high-quality corporate equities, sovereign debt of stable countries, real estate in growing regions). Some things hold value for years (growth equities with defensible moats). Some things hold value for quarters (speculative tokens, momentum trades).
Money needed in three months shouldn't be in assets with year-scale decay. Money you won't need for forty years can include assets with faster decay as long as they're appropriately sized.
Foundation
40-60 yrCash, sovereign debt, nuclear, grid, commodity production. Assets with multi-decade physical moats that cannot be replicated quickly.
Backbone
15-30 yrHard assets: gold, productive land, utilities, energy transport, logistics. Long-cycle assets with regulatory and capital barriers to entry.
Engine
5-10 yrDominant productive businesses. Semiconductors, compute hardware. The processing layer that powers everything above it.
Data Infrastructure
7-12 yrDiversified equity exposures. Cybersecurity, exchanges, payments. The data and transaction layer with strong switching costs and regulatory moats.
Interface
3-5 yrGrowth equities with identifiable moats. Enterprise SaaS, applications. Subject to AI disruption screening — seat-based pricing and low connectivity increase fragility.
Frontier
1-3 yrHigh-yield credit, emerging markets, thematic exposures, biotech, early innovation. High potential return, high decay rate. Requires strong conviction and tight position sizing.
Catalyst
tacticalSpeculative positions: venture equity, crypto tokens, single-stock concentrated bets, event-driven hedges. Not held for durability — held for a specific catalyst with defined exit.
Portfolio construction begins with a target allocation across layers — set by the client's goals, not by market conditions — and adjusts the layer weights based on regime. In a Hard Asset regime, Layer 2 weight increases. In a Growth regime, Layers 3–5 carry more weight. The structure is stable across regimes; the weights adapt.
The durability layer classification is implemented in nexus-core under
pw_classify_layer and
pw_get_layer_caps.
Individual holdings are classified programmatically; classification rules are public.
Question 3
Is this asset actually good of its kind?
Regime and durability tell us what kind of asset to hold. Quality scoring tells us whether this specific asset is a good instance of its kind. A holding can be in the right regime and the right layer and still be a worse choice than a peer.
Cash Generation (CROIC)
Cash return on invested capital — does the business actually generate cash per dollar of capital?
Financial Health
Piotroski F-Score — a 9-point financial health score covering profitability, leverage/liquidity, and operating efficiency.
Trend Persistence
Hurst exponent on the price series — is the series mean-reverting, trending, or neither? Informs position sizing and rebalancing cadence.
Perez Cycle
Where is the sector in the technological revolution framework? Early phases have different risk/return profiles than mature phases.
Lambda (λ) Decay
Decay constant — how fast does this asset's characteristic life decay? Informs expected holding period and rebalance costs.
Regime Compatibility
Does the current regime favor or disfavor this asset class? Applied as a weighting modifier.
Sector Composition
Within sector, what's the concentration? Diversification within a thematic exposure?
ASAN Screen
Is this exposed to AI disruption risk? Low-connectivity CRUD software, seat-based pricing, discretionary-spend categories vulnerable to AI-driven efficiency.
The scoring is transparent — every check cites its source and shows its calculation. A low score on a single check doesn't automatically rule out an investment, but a low score across multiple checks is a strong signal to look for alternatives.
The 8-check system is implemented in nexus-core under
pw_score_ticker.
Scoring results for any publicly traded security are reproducible; the data sources (SEC EDGAR,
market data vendors) are documented in the code.
Confidence Tiers
High Confidence
Full allocation within layer weight
Moderate Confidence
Reduced position, closer monitoring
Low Confidence
Minimal allocation, watch list only
Below Threshold
Excluded from new allocation
Putting It Together
How the three questions combine
For every holding in a client's portfolio, we can answer:
- What regime are we in, and does that favor or disfavor this asset class?
- What durability layer is this asset in, and does the client's target allocation support this exposure?
- What's the quality score, and how does it compare to alternatives in the same layer?
A holding that passes all three checks has a clear rationale. A holding that fails one or more has a clear reason to be adjusted, replaced, or exited.
Over time, this produces portfolios that are explainable at the individual-position level. When a client asks "why do I own this?" — or, more importantly, when a regulator or executor asks that question after the fact — the answer is documented in the scoring record.
Boundaries
What we don't do
Several things we don't include in the framework are worth stating explicitly.
We do not time markets on the basis of macro forecasts.
Regime detection is about identifying the current state, not predicting changes. If a regime change occurs, we update our allocation weights. We don't pre-position based on a prediction.
We do not use technical indicators as primary decision signals.
The Hurst exponent in the 8-check framework informs position sizing and rebalance cadence; it doesn't generate buy/sell signals on its own.
We do not treat Monte Carlo simulations as forecasts.
When we run Monte Carlo retirement scenarios, the output is a probability distribution of outcomes given our assumptions — not a prediction of what will happen. The assumptions are stated; changing them changes the output; the client sees both.
We do not use individual stock selection as a primary source of expected return.
The 8-check scoring is primarily a quality filter and risk-management tool. Our expected return comes from asset class exposure and regime compatibility, not from stock picking.
We do not recommend specific tax strategies.
Tax is a complex area requiring qualified professional advice. We work with our clients' tax professionals and coordinate as needed, but we don't provide tax advice ourselves.
Implementation
Open source and transparent
The analytical framework described above is implemented in open-source code at github.com/Protocol-Wealth/nexus-core under Apache 2.0 license. Anyone — clients, peer advisors, regulators, researchers — can read the code, verify the logic, and run the tools themselves.
We chose to open-source the analytical framework because we believe the right boundary between public and proprietary is not "how we score assets" but "what we recommend to specific clients given their circumstances." The framework is the method. The advisory service is the method applied to the particular client.
Academic sources for the components (Perez cycles, Piotroski scoring, Hurst exponent methodology, portfolio construction theory) are compiled at pwos.app/references.
Research Foundation
Intellectual Lineage
Every component of our process draws from established academic and practitioner research. We built the integration layer — the system that combines these inputs, applies them across traditional and digital assets, and adjusts allocation based on what the signals say.
| Component | Source | What We Do With It |
|---|---|---|
| Financial Health | Piotroski F-Score (Journal of Accounting Research, 2000) | Run it programmatically as a survival filter |
| Cash Generation | Greenblatt / quality-investing tradition | Apply as a binary gate for capital allocation |
| Trend Persistence | H.E. Hurst (1951); Mandelbrot applied to markets (1960s-70s) | Compute it as a momentum confirmation signal |
| Durability / Decay | Buffett moat framework; Mauboussin competitive advantage period | Formalize decay rates into a scoring input |
| Regime Detection | Hamilton regime-switching model (1989); Dalio economic machine | Specify which signals map to which regimes; maintain systematic classification |
| Technology Cycle | Carlota Perez, Technological Revolutions and Financial Capital (2002) | Classify where assets sit in the long-term technology adoption cycle |
| AI Disruption Screen | Christensen disruption theory applied to enterprise software | Screen SaaS for vulnerability: low connectivity + seat-based pricing |
| Layer Classification | OSI model (networking); Vaclav Smil (energy systems) | Apply infrastructure stacking to investment classification |
| Fragility / Antifragility | Nassim Nicholas Taleb, Antifragile (2012) | Stress-test positioning: what benefits from disorder? |
| Credit Cycle | Hyman Minsky, stability-instability hypothesis | Monitor credit conditions as a leading regime indicator |
| Skill vs. Luck | Michael Mauboussin, The Success Equation (2012) | Distinguish repeatable process from randomness in outcomes |
| Compounding Patience | Warren Buffett / Berkshire Hathaway capital allocation | Prefer durable compounders over high-turnover strategies |
| Energy Foundations | Vaclav Smil, Energy and Civilization (2017) | Ground durability model in physical energy constraints |
| Resource Constraints | Nicholas Georgescu-Roegen, entropy economics (1971) | Recognize that all economic value degrades — measure the rate |
Next
Go deeper
Framework
The broader context this investing approach sits within.
nexus-core on GitHub
The open-source implementation.
pwos.app/references
Academic sources for every framework component.
Security and infrastructure
How client investment data is protected.
RIA agent substrate
How the substrate behind this analysis enables agent training in an RIA context.
Last updated: June 5, 2026.
Past performance is not indicative of future results. This page describes our analytical framework; specific investment recommendations depend on individual client circumstances and are made through our advisory relationship. Protocol Wealth LLC is a SEC-registered investment adviser (CRD #335298).
Framework Disclaimer: The Protocol Wealth Asset Framework, built on the Entropic Macro Framework methodology — including the 7-layer durability model, 8-check scoring system, and related analytical methodologies — is a systematic framework built on established research. Framework scores, tiers, and classifications reflect historical and current quantitative metrics only; they do not constitute buy, sell, or hold recommendations for any specific security.
Educational Purpose: This content is for educational and informational purposes only. It does not constitute investment advice, a recommendation to buy, sell, or hold any security, or an offer or solicitation of any kind.
Risk Disclosure: All investments involve risk, including the potential loss of principal. Digital assets are highly speculative and volatile. Past performance does not guarantee future results.
Registration with the SEC does not imply a certain level of skill or training.