4. Where to go next: the bridge to the other tracks

Macro 6: From macro to strategy

This is the end of the macro course, and the right place for an honest stocktake of what you now have, what you do not have yet and where to get hold of what is missing.

What you have after six levels

• The price of money and discounting (level 1): why the risk-free rate is the bar everything else competes against, why a distant cash flow gets wrecked when that bar rises, why nominal figures mislead and what duration means.
• Who sets that price (level 2): how central banks work, what balance sheet expansion and contraction really do, and why a central banker's words move more than their decisions.
• The cycle and the curve (level 3): what information the shape of the yield curve carries, which phase you are in and, above all, that the market responds to the surprise and not to the data.
• Inflation and real assets (level 4): where it comes from, what protects you from it and what does not, and how currencies and commodities fit into the picture.
• Debt and credit (level 5): how leverage turns an orderly fall into a cascade and why correlations go to one exactly when diversification was the plan.
• And this level: regimes, exposure instead of forecasting, and execution conditions.

Notice what does not appear on that list: any ability to predict anything. That outcome is not a failure of the course, it is its purpose. What you have gained is a framework for asking, of any investment idea, which environment it works in and what would have to happen for it to stop working.

What you do not have yet

With what you have now you can reason well and be wrong elegantly, but you still cannot:

• Decide whether a difference in results between two strategies is real or fits comfortably inside chance.
• Tell luck from skill in a track record of returns.
• Build a historical test that does not lie to you.
• Source, clean and align the data needed to check any of this for yourself.

Those four things are, precisely, the content of the two tracks that follow.

The Intermediate track: math and statistics

This is not math for its own sake: it is the language in which the ideas of this course become measurable.

• Distributions, mean and dispersion, fat tails. 'The market moved a lot' becomes a comparable figure, and you find out why the Gaussian bell curve always underestimates episodes like the one on August 5, 2024.
• Correlation, and why it is not causation. Level 5 taught you that correlations go to one; here you learn to measure it and to see how it shifts depending on the period you pick.
• Significance and multiple testing. Module 1's warning about the four hundred combinations becomes a concrete penalty you can actually compute. It is López de Prado's central contribution to the overfitting problem.
• Out-of-sample validation. 'A strategy validated inside a single regime is not validated' stops being a phrase and becomes a procedure: split the sample, walk forward, do not look back.
• Regression. This is where 'knowing your exposure' goes from intuition to a number: you can estimate how much your portfolio responds to rates, to credit or to volatility, instead of having opinions about it.

The Technical track: Python

The Technical track exists because none of the above can be checked by hand. It gives you the ability to download data, tidy it up, build a backtesting engine and, above all, avoid the two traps that ruin most amateur backtests: look-ahead bias, unwittingly using information that was not published on that date, and survivorship bias, testing on a list of companies that leaves out the ones that went bust. There the calendar from this module stops being a warning and becomes just another column in your dataset, and execution costs stop being a vague worry and turn into a parameter you can vary until you see at what level your strategy stops making sense.

Long-Term Capital Management: the case that sums up the course

The fund was set up in 1994 with the most qualified team ever assembled: top-tier traders and two future Nobel laureates in Economics among its partners. In its early years it delivered high double-digit annual returns with apparently low volatility, resting on enormous leverage and on models that estimated the normal behavior of spreads with precision.

In 1998 Russia defaulted on its debt and the market stopped behaving normally. Positions that looked independent of one another turned out to be the same bet repeated: that spreads narrow and that there will always be somebody on the other side to close out with. The fund's capital fell from around $4.7 billion to some $400 million in the space of four months, and the Federal Reserve Bank of New York had to coordinate a rescue of around $3.6 billion put up by fourteen institutions to avoid a disorderly liquidation.

They were not short of mathematics. What they were short of was having written down which exposure all their positions really shared, and which regime was needed for their models to carry on describing the world. It is the whole course summed up in a single case.

A warning that is not a formality

All this material is education, not financial advice. Nothing you have read is a recommendation to buy or sell any asset, nor is it tailored to your circumstances, your objectives or your tolerance for risk. The historical cases are used to illustrate mechanisms, and the fact that something happened once does not mean it will happen again. Investing carries the risk of loss, including the total loss of your capital. If you are going to take decisions with real money, do it on your own analysis and, if you need it, with a professional authorized to provide it.