Your Risk Model Was Built for a Different Economy. Here Is the Math.
Every Deal Answers One Question.
Tamika Tyson, Founder & Managing Partner, SCALE
April 2026 | Part 7 of 7 in the Through the Cycle series
Last week I argued that in a world where everyone has AI, the edge comes from better thinking. I showed you where artificial intelligence excels, where it breaks, and where humans fail with it. I gave you seven principles for keeping your thinking ahead of your tools. But a framework without measurement is philosophy. This is the post where we turn philosophy into practice. There is more math in this post than in the other six combined. That is the point.
Every risk professional has had this experience. You are sitting in a room. The deal team is presenting the best deal they have ever seen. The game-changing deal. The numbers are compelling. The story is tight. The excitement is real. Then you go back to your desk. You dig into the financials, the risk profile, the industry, the structure. And you think: why does the deal team believe this is the best deal ever?
This happens all the time. It is unavoidable. And here is the thing: sometimes they are right. Some of those deals generate outsized returns. But they require real structuring and deep knowledge to get there. You have to dig in, identify the risks, structure around the ones you find unacceptable, and price the ones you choose to accept. I have done this more times than I can count. And in every single case, I ask the same question: what is the next best alternative?
That question is the thread that runs through this entire series. When you think about the cycle position, the political variable, the credit infrastructure, the investment cycle stage, the deal itself, the people involved, and all of the tools we have discussed across seven posts, the question that sits on top of everything is: given all of this, what is the next best alternative? Because risk is not just about whether this deal works. It is about whether this deal is the best use of capital given everything else you could do with it, in this cycle, at this stage.
Over the last six weeks, we have walked through the machinery together. Behavioral discipline when markets turn (Post 1). Risk frameworks and the math behind uncertainty (Post 2). Credit infrastructure and where the plumbing breaks (Post 3). The political variable your risk model is probably ignoring (Post 4). The twelve-stage Investment Cycle and fifty-five years of evidence (Post 5). Critical thinking as the only real edge in an AI-saturated world (Post 6). Now, in this final piece, we arrive at the practical mechanics: how do you actually measure it.
The Paradox That Kills
Let me start with the observation that has shaped my entire career: a risk model built during expansion will be validated by the very conditions it was designed to measure. Then the cycle turns.
This applies to credit models and investment models alike. Think about what happens during an expansion. You build a model. Maybe it is a credit scoring model, maybe it is a DCF, maybe it is a portfolio allocation framework driven by historical returns. You backtest it against the last five years of data. The accuracy is 94 percent. The model is approved. You deploy it into production. For three more years, it continues to perform beautifully. Everything you predicted happens. The model is validated. Your confidence grows. You load more capital into the positions it recommends.
Then, in year six, something shifts. Borrowers default earlier than the credit model said they would. The equity multiples your valuation model relied on compress. Recovery rates are worse than historical averages. The correlation assumptions in your portfolio model, which looked conservative during calm markets, turn out to be meaningless in stress. The model breaks not because it was poorly constructed, but because it was built during a cycle that had a specific character, and now the cycle has a different character. Your risk model was built for a different economy.
This is the foundational problem of risk measurement. Every risk model, whether it measures credit losses or investment returns, is a compressed version of recent history. When the recent history changes, the model becomes dangerous precisely because you trust it most. That is the paradox. And this is where the question matters most: if the model is telling you this is a good deal, but the cycle has shifted, what is the next best alternative?
The practitioners who understand this run two models. They do not think of it as a choice between them. They think of it as answering two different questions.
The Two Questions
Through-the-Cycle (TTC) asks: what is the long-term average risk of this exposure, given all the different economic regimes in the historical data? TTC is stable. It does not move much from quarter to quarter. It is appropriate for structural decisions: how much capital to hold, how much illiquidity to absorb, how to size your balance sheet or your fund for the long run.
Point-in-Time (PIT) asks: what is the current risk of this exposure, given the economic conditions right now? PIT is responsive. It moves when conditions change. It is appropriate for operational decisions: whether to deploy, whether to hedge, whether to take risk off the table, whether this deal is worth doing today.
The mistake practitioners make, again and again, is thinking they can choose one or the other. They cannot. You need both. TTC for the decisions that have to survive a full cycle. PIT for the decisions that have to respond to where the cycle is right now.
This dual lens applies to all three cycles we have discussed in this series, not just credit. The investment cycle has a TTC view (the long-run return profile across regimes) and a PIT view (what stage are we in right now, and what does that mean for deployment). The credit cycle has a TTC view (average default rates and loss severity across a full cycle) and a PIT view (current spreads, current stress, current conditions). The political cycle has a TTC view (over a full cycle, political intervention amplifies both the boom and the bust) and a PIT view (what is the current administration doing right now to accelerate or restrain the other two cycles).
| Cycle | TTC View (Structural) | PIT View (Operational) |
|---|---|---|
| Investment Cycle (the "What") | Long-run return profile across regimes. How should you size your portfolio for the full cycle? | Which of the twelve stages are we in? Is the trajectory improving or deteriorating? Should you deploy or preserve? |
| Credit Cycle (the "Why") | Average default rates, loss severity, and recovery rates across a full cycle. What capital buffer do you need? | Current spreads, current default trajectory, current borrower stress. What is the real risk right now? |
| Political Cycle (the "Who") | Over a full cycle, policy amplifies both the boom and the bust. Build for oscillation. | What is current policy doing to the other two cycles? Easing, tightening, disrupting, deregulating? |
I have seen organizations fail because they were running pure TTC and did not see the turn coming until it was too late to act. I have also seen organizations fail because they were running pure PIT and when the turn came, their capital was gone, deployed into risk-on positioning that looked attractive under the prior regime. In both cases, the organization never asked the question: given where we are in the cycle, what is the next best alternative to the position we are holding?
The Metrics That Matter
Each cycle has its own set of compounding risk metrics. Understanding how they move together, within each cycle and across all three, is the heart of practical risk management.
The Credit Cycle
Three components drive credit risk. Probability of Default (PD) is the first. During a normal cycle, a BB-rated obligor might have a PD of 1-2 percent. In a severe recession, that same obligor might have a PD of 15 percent or higher. The rating did not change. The economic conditions shifted the distribution of outcomes.
Loss Given Default (LGD) is the second, and it is where practitioners make their most dangerous mistake. They treat LGD as a constant. They say it is 45 percent and never revisit it. During expansions, when asset values are rising and liquidity is abundant, LGD ranges from 30 to 40 percent. During recessions, when asset values are falling and liquidity evaporates, LGD ranges from 60 to 75 percent or higher. The collateral that was worth $10 million on your balance sheet in 2005 was worth $4 million in 2009. Recovery rates collapsed. LGD exploded. If you were using a static LGD assumption, you were underestimating your losses by a factor of two.
Exposure at Default (EAD) is the third. On its surface, it seems straightforward: if you have loaned someone $10 million, your EAD is $10 million. But credit works through more complex structures. You have committed lines, letters of credit, guarantees. When a borrower gets into distress, they draw on everything available. We call this the "race to default." Your EAD grows exactly when your appetite to hold exposure is shrinking.
These three multiply together: Expected Credit Loss (ECL) = PD x LGD x EAD. What makes this calculation treacherous is that all three move together, in the direction that is worst for you. When growth slows, PD goes up. Asset values fall, so LGD goes up. Stressed borrowers draw, so EAD goes up. The losses multiply.
The Investment Cycle
Three parallel components drive investment risk. Earnings Deterioration Risk is the first. This is the probability that returns fall short of what the model projected. During expansion, earnings grow steadily and your model looks accurate. During contraction, earnings decline faster than projected, and the gap between expected and actual returns widens. The forward P/E of 20x that looked reasonable at Stage 5 of the Investment Cycle is dangerous at Stage 9, because the earnings it is based on are about to contract.
Valuation Compression is the second. This is the investment cycle's version of LGD. When the cycle turns, multiples compress. The asset that was valued at 8x EBITDA in a growth environment gets repriced to 5x in a contraction. Your portfolio's mark-to-market value falls not because the businesses stopped operating, but because the market's willingness to pay for future earnings has changed. If you are holding illiquid positions, the compression is even worse because you cannot exit at any price.
Deployment Exposure is the third. This is the investment cycle's version of EAD. How much of your available capital is deployed, and how does that compare to the opportunity set? In late expansion, deployment rates peak because everything looks attractive. When the cycle turns, you discover you are fully invested with no dry powder, precisely when the best opportunities are about to emerge. The capital is locked up when flexibility matters most.
These three compound the same way: when the cycle turns, earnings deteriorate (returns fall short), multiples compress (values drop), and you are over-deployed (exposure is too high). All three move against you simultaneously.
The Political Cycle
Three components drive political risk, and they act as accelerants on the other two cycles. Policy Rate vs. Neutral Rate is the first. The gap between the Fed funds rate and the estimated neutral rate tells you whether monetary policy is adding fuel or applying brakes. When the policy rate sits well below neutral, money is cheap, credit expands, and the investment cycle accelerates. When it sits above neutral, the opposite. This is the most direct transmission mechanism from political decisions to the other two cycles.
Fiscal Impulse is the second. This measures the change in the government's fiscal stance: not the deficit itself, but whether the deficit is growing or shrinking relative to GDP. A positive fiscal impulse means the government is injecting stimulus. A negative one means it is withdrawing support. A government running a large positive fiscal impulse during late expansion is pouring gasoline on a fire. A government withdrawing fiscal support during early contraction is pulling the safety net when people are starting to fall.
Regulatory and Trade Exposure is the third. Tariffs, sanctions, sector-specific regulation, deregulatory shifts: these change the rules mid-game. Your portfolio's exposure to policy-sensitive sectors determines how much political risk you are actually carrying. This is the component most practitioners underweight, and it is especially relevant right now.
These three compound the same way: when monetary policy tightens while fiscal deficits are unsustainable and trade policy is disrupting supply chains, all three move against you simultaneously. The political cycle does not just set the backdrop. It changes the math on the other two cycles.
The Composite Dashboard: April 2026
I have constructed a composite indicator that combines nine key metrics, grouped by cycle. Each metric is normalized on a scale from -1 (severe contraction) to +1 (strong expansion) using z-scores against their own historical distributions, then averaged equally within each cycle sub-group. The three sub-scores are weighted equally to produce the composite. Equal weighting is a deliberate choice: it prevents any single cycle from dominating the signal and forces the practitioner to examine the breakdown rather than relying on the headline number.
Investment Cycle indicators: yield curve slope, unemployment rate, PMI. Current sub-score: -0.45. The economy is slowing but not yet contracting. Leading indicators are turning. The yield curve has been inverted for months.
Credit Cycle indicators: credit spreads, the Baltic Dry Index (a real-time measure of global trade demand), Debt-to-GDP, default rates. Current sub-score: -0.65. This is the weakest of the three. Spreads are widening, trade volumes are declining, and default rates are beginning to accelerate. The credit cycle is ahead of the investment cycle in its deterioration.
Political Cycle indicators: ISM (as a proxy for policy transmission), the VIX (which measures implied volatility and spikes when fear reprices risk). Current sub-score: -0.60. Policy uncertainty is elevated. Tariff activity is disrupting supply chains. The fiscal impulse is contracting.
The composite score across all nine is -0.57. This reads as late contraction. Not yet crisis conditions. But the breakdown tells the real story: credit is deteriorating fastest, the political variable is adding friction, and the investment cycle is starting to follow. The investment cycle sub-score sits in early contraction territory, which means it has not yet caught up with the other two. That lag is not reassuring. It means the worst of the investment cycle deterioration may still be ahead. You manage each cycle differently, which is why the breakdown matters more than the single number.
What does this mean for risk parameters? The practitioner applies a regime multiplier:
| Regime | Composite Score | PD Multiplier | LGD Multiplier | Recovery Multiplier |
|---|---|---|---|---|
| Expansion | 0 to +1.0 | 0.7 - 0.9 | 0.9 - 1.0 | 1.0 - 1.1 |
| Late expansion | -0.25 to 0 | 0.95 - 1.0 | 1.0 - 1.05 | 1.0 |
| Early contraction | -0.50 to -0.25 | 1.2 - 1.4 | 1.10 - 1.20 | 0.85 - 0.95 |
| Late contraction (current) | -0.75 to -0.50 | 1.6 - 2.0 | 1.30 - 1.50 | 0.70 - 0.85 |
| Crisis | Below -0.75 | 2.0 - 3.0 | 1.50 - 2.0 | 0.50 - 0.70 |
These multipliers reflect what actually happens to risk across different regimes. I have derived these ranges from observed default rates, recovery outcomes, and valuation data spanning the four complete post-Volcker cycles (1982-2000, 2000-2008, 2008-2019, 2020-2023), calibrated against Moody's annual default studies and the Basel Committee's downturn LGD research. They are not theoretical. They are observed.
The Case Study: One Company, Three Cycles
Let me make this concrete. A mid-market manufacturing company. $200 million in revenue. BB-equivalent credit profile. It supplies industrial components, sources 40 percent of its total inputs from tariff-affected countries, and has $75 million in outstanding debt with a $25 million undrawn revolver. This company could be a borrower in your loan book, a holding in your fund, a counterparty in your supply chain, or a portfolio company in your family office. The math works the same regardless of where you sit.
Through the credit cycle lens. Under baseline TTC assumptions, the company's PD is 1.5 percent, LGD is 40 percent, and EAD is $100 million (the $75 million drawn plus the full $25 million revolver, because in a stress the company will draw everything available). Expected credit loss: $600,000. Manageable.
Now apply the late contraction multipliers. PD moves to 2.7 percent (1.5 x 1.8). LGD moves to 56 percent (40 x 1.4). The company has already drawn $10 million of the revolver, so EAD is now $100 million. Expected credit loss under stress: $1,512,000. That is 2.5 times the baseline. And this is only one lens.
Through the investment cycle lens. Six months ago, this company was valued at 7x EBITDA. EBITDA was $30 million. Enterprise value: $210 million. Today, the composite is in late contraction, though the investment cycle itself is lagging the other two. Earnings have softened. EBITDA is trending toward $25 million as orders slow. And the market multiple for industrial manufacturers has compressed from 7x to 5x. New implied enterprise value: $125 million. That is a 40 percent decline in value in six months, and the investment cycle has not yet caught up with the credit and political deterioration. If you are an investor holding equity in this company, your mark-to-market loss is $85 million. If you are a lender, your collateral coverage just deteriorated significantly.
Through the political cycle lens. The company sources 40 percent of its total inputs from countries subject to new tariffs. On a $200 million revenue base, that is roughly $80 million in tariff-exposed procurement. The tariff rate increased from 5 percent to 25 percent over the last year. That adds $16 million to annual input costs. The company cannot pass all of that to customers because its domestic competitors source differently. Margins compress by 3-4 percentage points. The EBITDA decline is not just cyclical. It is structural, driven by a political decision that changes the company's cost base for as long as the tariffs remain in place. The fiscal impulse is contracting, so there is no offsetting stimulus. And the policy rate is above neutral, which means borrowing costs on that $75 million in debt are higher than they were a year ago.
The three cycles converge. The credit lens says expected losses have more than doubled. The investment lens says the company's value has dropped by 40 percent. The political lens says margins are structurally impaired and borrowing costs are elevated. None of these is catastrophic in isolation. Together, they paint a picture of a company under compounding stress from three directions simultaneously. The credit model alone would have told you the risk was elevated but manageable. The full picture tells you something different.
Now multiply this across a hundred obligors in a $500 million portfolio, and you understand why the composite dashboard matters. The question is not whether any single company survives. The question is: given everything you now see across all three cycles, what is the next best alternative?
The Pitfalls
There are eight mistakes I see practitioners make repeatedly when they try to measure risk through the cycle:
| Pitfall | Why It Matters |
|---|---|
| 1. Relying on historical correlations | Correlations are unstable. In normal times, two assets may be uncorrelated. In a crisis, everything correlates toward one. Historical correlations measured during calm periods are useless in stress. This applies across all three cycles: credit spreads, equity returns, and policy responses all converge in a downturn. |
| 2. Ignoring regime shifts | A 30-year average default rate of 1.2% might mask 15 years at 0.5% and 15 years at 1.9%. The same applies to equity valuations and policy environments. The average hides the volatility between regimes. |
| 3. Treating loss severity as static | In credit, LGD can double or triple when the cycle turns. In investment portfolios, valuation compression accelerates in the same conditions. In the political cycle, fiscal capacity to absorb shocks erodes precisely when it is needed most. None of these are constants. All of them move against you at the worst time. |
| 4. Undercounting exposure | Credit practitioners focus on PD and LGD and miss that borrowers draw on every available line in a stress. Investment practitioners miss that deployment rates peak right before the cycle turns. And almost everyone underweights how much of their portfolio is exposed to political and regulatory risk. Exposure grows across all three cycles when you least want it to. |
| 5. Confusing correlation with causation | Unemployment and defaults both rise in a downturn, but neither causes the other. Tariffs and inflation both rise in a trade war, but neither causes the other. They are symptoms of the same underlying stress. False attribution leads to false risk drivers. |
| 6. Updating the model too frequently | Refitting every quarter introduces noise. You chase the last data point instead of understanding the signal. Annual updates aligned with the cycle are appropriate. Quarterly is too frequent. Monthly is gambling. |
| 7. Failing to stress-test the unexpected | Your model assumes rates do not spike suddenly, employment does not drop 10% in a quarter, liquidity does not vanish, and trade policy does not reverse overnight. In a true stress, all of these happen simultaneously. Test for them. |
| 8. Ignoring the political variable | This is the most common blind spot. Practitioners build sophisticated credit and investment models and treat the political cycle as background noise. It is not background noise. It is the accelerant. Policy rate decisions, fiscal expansion, tariffs, sanctions, and regulatory shifts do not just create headlines. They change the math on the other two cycles. If your model does not have a political input, your model is incomplete. |
The Practitioner Framework: Where Are We in the Cycle?
Given all of this, how should you organize your risk management? The framework below is designed to answer one question first: where are we in the cycle? Because until you know where you are, you cannot answer the question that matters: what is the next best alternative?
| Pitfall | Why It Matters |
|---|---|
| 1. Relying on historical correlations | Correlations are unstable. In normal times, two assets may be uncorrelated. In a crisis, everything correlates toward one. Historical correlations measured during calm periods are useless in stress. This applies across all three cycles: credit spreads, equity returns, and policy responses all converge in a downturn. |
| 2. Ignoring regime shifts | A 30-year average default rate of 1.2% might mask 15 years at 0.5% and 15 years at 1.9%. The same applies to equity valuations and policy environments. The average hides the volatility between regimes. |
| 3. Treating loss severity as static | In credit, LGD can double or triple when the cycle turns. In investment portfolios, valuation compression accelerates in the same conditions. In the political cycle, fiscal capacity to absorb shocks erodes precisely when it is needed most. None of these are constants. All of them move against you at the worst time. |
| 4. Undercounting exposure | Credit practitioners focus on PD and LGD and miss that borrowers draw on every available line in a stress. Investment practitioners miss that deployment rates peak right before the cycle turns. And almost everyone underweights how much of their portfolio is exposed to political and regulatory risk. Exposure grows across all three cycles when you least want it to. |
| 5. Confusing correlation with causation | Unemployment and defaults both rise in a downturn, but neither causes the other. Tariffs and inflation both rise in a trade war, but neither causes the other. They are symptoms of the same underlying stress. False attribution leads to false risk drivers. |
| 6. Updating the model too frequently | Refitting every quarter introduces noise. You chase the last data point instead of understanding the signal. Annual updates aligned with the cycle are appropriate. Quarterly is too frequent. Monthly is gambling. |
| 7. Failing to stress-test the unexpected | Your model assumes rates do not spike suddenly, employment does not drop 10% in a quarter, liquidity does not vanish, and trade policy does not reverse overnight. In a true stress, all of these happen simultaneously. Test for them. |
| 8. Ignoring the political variable | This is the most common blind spot. Practitioners build sophisticated credit and investment models and treat the political cycle as background noise. It is not background noise. It is the accelerant. Policy rate decisions, fiscal expansion, tariffs, sanctions, and regulatory shifts do not just create headlines. They change the math on the other two cycles. If your model does not have a political input, your model is incomplete. |
I am deliberately not giving you a formula for answering that question. The question is a forcing function, not a calculation. It exists to make you stop and think before you act. If I gave you a model that produced a ranked list of alternatives, you would do what practitioners always do with models: read the output and stop thinking. That is the opposite of what this series is about.
There is something else the question forces you to confront. Doing nothing is an alternative. It is almost never on the list. In my experience, the deals that cause the most damage are not the ones where the analysis was wrong. They are the ones where a senior leader wanted the deal done, and the analysis was shaped to support a conclusion that had already been reached. I call these FOCs: Friends of the C-Suite. Everyone in the room knows the deal is marginal. Everyone in the room knows the risk parameters are being stretched. But the deal has a sponsor, and the sponsor has gravity, and the gravity pulls the analysis toward yes. The question "what is the next best alternative?" is the only thing that can break that gravity, because it forces the room to compare the deal on the table against every other use of that capital, including the use where you do not deploy at all. Doing nothing is a decision. It should always be on the table. It rarely is.
And sometimes the deal is the deal. You have done the work. You understand the risks. The structure is not perfect, the timing is not ideal, and the cycle position adds complexity you would rather not carry. But the deal is what it is, and the honest assessment is that it clears the bar. Not every marginal situation is a FOC. Sometimes the right move is to acknowledge the imperfections, price them in, and proceed. The discipline is knowing the difference between a deal that is genuinely marginal and a deal that is being pushed past marginal by someone with gravity.
These six steps are the baseline of professional risk management. Organizations that do all six survive the cycle. Organizations that skip steps pay.
As I look back across these seven posts, I see a narrative arc.
PostTitleKey Insight1The Cost of PanicPanic selling costs more than any market crash. Behavioral discipline is the foundation.2Risk FrameworksThe math behind uncertainty. If you watch the right indicators, you know something is wrong before the headlines confirm it.3Credit InfrastructureMoney does not make the world go round. Credit does. Where the plumbing breaks is where systemic risk lives.4The Political VariablePolitical decisions are the mechanism through which credit and investment cycles accelerate, pause, and reverse.5Three Cycles, One EconomyThe twelve-stage Investment Cycle from 1971 to today. The same five steps preceded every major crisis.6Critical Thinking Equals AlphaThe edge belongs to the people who think critically. AI sees the straightaway. The cycle turns at the corners.7The MathThree cycles. Three sets of compounding metrics. One question: what is the next best alternative?
All seven pieces are necessary. If you understand the cycle but cannot measure it, you are still flying blind. If you can measure it but do not understand the pattern, your measurement is mechanical and disconnected from reality. If you have the metrics and the pattern but lack the discipline to execute quarterly and update monthly, you will fail when it matters most.
Closing
The composite indicator is currently at -0.57. This is late contraction. It is not a prediction of a crisis tomorrow. It is a statement that the regime has shifted. Credit is leading the deterioration. The political variable is adding friction. The investment cycle is starting to follow.
If you are deploying capital, whether as a lender, a fund manager, or a family office, this is the moment to tighten. Not to panic. Not to exit every position. But to be selective, to reduce concentration, to ensure your capital is positioned against scenarios where conditions worsen rather than improve.
If you are managing a business, whether as a corporate treasurer, a CFO, or a founder, this is the moment to shorten duration on your debt, to lock in rates if you have not already, to build liquidity buffers, to think carefully about leverage.
If you are setting policy, whether as a regulator, a board member, or a risk committee chair, this is the moment to resist the impulse to ease every time conditions turn soft. The organizations that survived 2008 did so because they had buffers. The organizations that broke did so because they had spent their buffers during the prior expansion.
But contraction is not all bad. When there is blood in the streets, there is money to be made. The practitioners who prepared during expansion, who maintained discipline, who kept dry powder, who ran the six steps when it was boring and the markets were calm, those are the practitioners who deploy into distressed assets at the bottom. The best vintages in private equity, the best entry points in credit, the best acquisitions in corporate history all happened during contractions. The cycle punishes the unprepared. It rewards the disciplined.
The cycle will turn. It always does. The capitalization will be different. The industries will be different. The cast of characters will change. But the pattern underneath is the same. And the question that matters, whether you are sitting in a deal room or running a portfolio or setting policy, is the same question it has always been: what is the next best alternative?
I am currently in conversations with CEOs, CRO, CCO and boards applying this three-cycle framework to their specific portfolios and governance mandates. Reach me directly at tamika@tamikatyson.com.
Sources
Through-the-Cycle (TTC) vs. Point-in-Time (PIT) modeling frameworks: Basel Committee on Banking Supervision, "Studies on the Validation of Internal Rating Systems," Working Paper No. 14, revised May 2005; Moody's Analytics, "Through-the-Cycle and Point-in-Time EDF Credit Measures," 2011.
BB-rated obligor PD ranges (1-2% normal cycle, up to 15%+ in recession): S&P Global Ratings, "2023 Annual Global Corporate Default and Rating Transition Study," 2024; Moody's Investors Service, "Annual Default Study," various years.
LGD variation (30-40% in expansion, 60-75%+ in recession): Moody's Analytics, "Ultimate Recovery Database"; Basel Committee on Banking Supervision, "Guidance on credit risk and accounting for expected credit losses," December 2015.
Yield curve inversion predicting recession (7 of 8 since 1968, 87.5% accuracy): Federal Reserve Bank of New York, "The Yield Curve as a Leading Indicator," various research notes; Estrella and Mishkin, "Predicting U.S. Recessions: Financial Variables as Leading Indicators," Review of Economics and Statistics, 1998.
Baltic Dry Index: Baltic Exchange, daily freight rate data; Bloomberg Terminal historical series.
VIX (CBOE Volatility Index): Chicago Board Options Exchange, "VIX Index Historical Data," cboe.com; Whaley, Robert E., "The Investor Fear Gauge," Journal of Portfolio Management, 2000.
Debt-to-GDP ratios: Bank for International Settlements, "Credit to the Non-Financial Sector" database, bis.org; Federal Reserve Bank of St. Louis, FRED Economic Data.
Credit spreads and default rate data: ICE BofA US High Yield Index Option-Adjusted Spread; S&P Global, "Global Corporate Default and Transition Study," annual editions.
Equity valuation multiples and compression data: Aswath Damodaran, "Equity Risk Premiums," Stern School of Business, NYU, annual updates; S&P Capital IQ, historical EBITDA multiple data by sector.
Fiscal impulse and policy rate data: Brookings Institution, "Hutchins Center Fiscal Impact Measure"; Federal Reserve Economic Data (FRED), federal funds rate and neutral rate estimates.
Trade policy and tariff impact: Peterson Institute for International Economics, tariff tracker and trade policy analysis; U.S. International Trade Commission, economic impact assessments.