Chapter 31: The full loop

One real holding, taken from ledger to thesis with everything you now know.

By Matt Lang and Bree Lang · About the authors

Where this is on Sifter: Portfolio + Research, end to end

Now that we have made it through the entire book, it is time to put all of the tools to work at the same time. In this chapter, we will take one real holding from your book and run it through the full loop, from the transaction ledger all the way to a finished recommendation. Choose the holding you would defend least confidently if a committee asked you about it tomorrow, because that is the one that needs the work. By the end, you will have a five-line mini-thesis that is short enough to deliver to an investment committee without notes, and every single line will cite a number that you computed or verified with your own hands.

Budget an hour for the loop. That may sound like a lot, but there is no new machinery anywhere in this chapter, and every surface you will visit below is one you have already used.

The five steps below produce those five lines in order, and each step opens the page it needs.

With line 1 in hand, the next step asks what the position has done for the book.

Lines 1 and 2 describe what you own and what it has done. Line 3 asks what it is worth, and it takes two methods.

Line 4 is the only line that comes from a scorecard, and it is the one you have to verify yourself before it counts.

The last step is about the book rather than the holding, and it produces a question rather than a line.

Try it on Sifter: Step 5: how the position fits in the fund (Chapter 9 and Chapter 21)

A thesis about one holding is not finished until it says something about the book that holds it. Open Holdings and read the position's weight against the fund's single-position cap (if the fund has adopted limits, the cap is on Risk Flags; if it has not, the proposed line is there instead). Open Analyze › Attribution and read its MCTE, the share of the fund's tracking error it is spending, next to that weight. Then open Analyze › Diagnostics and check whether it is inside a correlated cluster, and with whom. Write one sentence that puts those three facts side by side, and then write the one question they raise for the committee, rather than the answer. A holding that is 6% of the book, 30% of its active risk and one of four holdings in a cluster raises a different question from a holding that is 6%, 4% and alone, and either question belongs in the meeting.

Before you write your own line 5, test the reasoning on a made-up loop. The question below gives you the five numbers and asks for the action.

Check yourself

Your loop produces: CTE +0.9pp (top contributor), MCTE 35% of fund TE on an 8% weight, DCF fair value 5% above price, comps at a 20% premium to peers, earnings quality B. The most defensible line 5 is:

Add: it is the fund's best performer. Past contribution tells you what happened, not what comes next. Looking forward, the valuation is full on both methods (a DCF barely above price and a comps premium), and the position is carrying an outsized slice of the risk budget while it stays there.
Trim: the contribution is banked, the valuation is full, and the position consumes a third of the fund's active risk at an 8% weight. (correct) Right. Notice that every clause in this line leans on a chapter, since the CTE is banked (Chapter 8), the valuation is full on both methods (Chapter 16 and Chapter 17), and the MCTE is far above the capital share (Chapter 8 and Chapter 10). A B on quality is perfectly fine, but it is not strong enough to overturn the risk-for-value trade.
Exit entirely: a comps premium means overvalued. Not quite. A premium to peers combined with a DCF near the price argues that the position is fully valued rather than broken, and the B on quality with a strong contribution history argues against going all the way to zero. Remember that conviction sizing works like a dial, and trimming is comfortably between "add" and "exit".

Line 5 has one more requirement, and the second question asks why it is there.

Check yourself

Line 5 ends with "the one number that would change your mind." Why does the template demand it?

It makes the pitch sound rigorous to recruiters. It does, but only as a byproduct, because the real function of the falsifier is mechanical rather than cosmetic.
A thesis with a stated falsifier can be monitored, and one without can only be re-argued. (correct) Right. Take a look at what a line such as "Trim, unless FCF conversion recovers above 70% by Q3" actually does, which is turn next quarter's filing into a decision trigger instead of a fresh debate. A fund that writes falsifiers ends up turning its book over on evidence, and a fund that skips them ends up turning its book over on whatever the room happens to feel that week.
Committees require it formally. Few committees formally require it, but the good ones tend to behave as if they did.

Now write yours. The lab below is the deliverable, and it ends with a count you can check.

Lab: Final deliverable: the mini-thesis

Write the five lines in a workspace note (Workspace › New › Note), one line per template slot, and make sure each line cites at least one number you computed in steps 1 through 4. Add the question from step 5 under the five lines, with the three numbers that raised it.
Attach or paste the two artifacts that carry the most weight, meaning your DCF sensitivity row and the holding's attribution line.
Read the five lines out loud once, and if any line cannot survive the question "says who?", replace the adjective with the number.
Keep the note when you are finished, since it is the seed of a real pitch and it is the shape of the reasoning a recruiter reads in the Training Floor exercises.

One step remains after the note is written, and it is the one that connects this book to the Training Floor.

The platform will keep computing these numbers for you, just as it did before you knew what any of them meant. The difference, a whole book later, is in you, because you now know what each number is, where it comes from, and, most valuably of all, when to stop trusting it. Even the methodology document admits in its own fine print that the numbers have limits, and knowing where those limits are is what makes someone an analyst instead of a reader of dashboards. The loop you ran in this chapter was practiced on a single holding, but it works the same way on every holding you will ever cover, and we hope you keep running it long after you close this book.

The Analyst's Handbook is Sifter's textbook for student investors, written by Matt Lang and Bree Lang. Sifter is free for students. This chapter as markdown.