Budget as Percent of Liquid Net Worth

Position sizing by liquid net worth is the boring sentence that usually arrives after the story ends. Before that, people size by tip-group conviction, by how loud a chart feels, or by "I can average down." ConvexClub treats budget as a first-class field because leverage that is not named as a percent of what you can actually lose tends to rewrite itself under stress.

This is not a formula for how much capital to put on any idea. It is a process essay about why the denominator has to be liquid net worth, and why public blowups keep reminding us what happens when budget language is opaque.

In this guide

1. What "budget" means in a process journal
2. Why liquid net worth is the useful denominator
3. Archegos as a budget-process failure (not a how-to)
4. What retailers usually get wrong in notes
5. A past-tense budget packet you can reuse
6. Liquid NW vs "I can always raise"
7. Comparing experiences without vanity precision
8. What we are not saying
9. Worked reconstruction (labeled composite)
10. Try this yourself

What budget means in a process journal

On ConvexClub, budget is capital at risk against a stated max-loss shape—not "position size in shares," not notional on a swap screen you do not keep, and not "whatever the Discord sized." Related framing: paper trading vs lived budget risk and the five factors.

A usable budget line answers three questions in past tense:

• How much cash-like wealth could absorb a total loss without forcing fires elsewhere?
• What fraction of that liquid base did I actually put at risk?
• What max-loss shape did I claim—and did my instrument match that claim?

If you cannot answer those after the fact, you did not have a budget. You had a purchase.

Why liquid net worth is the useful denominator

Gross net worth is full of frozen stuff: house equity you will not sell at a discount next Tuesday, private stakes with unknown bids, options that look large on a statement and small in a margin call. Liquid net worth is the duller pile: cash, marginable securities you can actually sell, and anything else that turns into dollars fast enough to matter when the thesis slides.

Using total NW as the denominator flatters the story. Using "account equity in this brokerage" understates household risk if the same person carries concentrated bets elsewhere. Liquid NW is a compromise: strict enough to sting, honest enough to compare across experiences.

We do not claim a sacred percent. We claim an unidentified percent is how people discover leverage after the fact.

Archegos as a budget-process failure (not a how-to)

In March 2021, Archegos Capital Management—a family office—defaulted after large, highly leveraged equity exposures moved against it. Credit Suisse, one of the prime brokers, later commissioned Paul, Weiss, Rifkind, Wharton & Garrison to investigate how the bank lost approximately $5.5 billion from the relationship. That figure is Credit Suisse's own disclosed loss magnitude in the investigation materials filed with its Form 6-K / Exhibit 99.2 (29 July 2021), not a ConvexClub estimate.

The same report reconstructs the late-March broker call when Archegos characterized having on the order of $120 billion in gross exposure against roughly $9–$10 billion in equity. Those are the bank's investigative summary numbers about the client's stated leverage character—not a template for anyone reading this to replicate financing.

Reuters' contemporaneous wrap of the post-mortem emphasized the report's conclusion: a "fundamental failure of management and controls" in Credit Suisse's prime services / investment bank, including chronic limit breaches and weak collateral discipline. Again: that is bank process language. Your journal is not a prime brokerage.

Why put a multi-billion-dollar bank failure next to a retail liquid-NW field? Because the failure mode is recognizable at smaller scale. Budget language that cannot be stated as a fraction of survivable wealth becomes invisible leverage. Total-return swaps and margin are just louder versions of "I sized it by how sure I felt."

We are not teaching swap structure. We are saying: if the Paul, Weiss report needed 165 pages to explain how opacity and overrun limits became a $5.5bn hole for one bank, your one-line note that says "concentrated, high conviction" is not a size control.

What retailers usually get wrong in notes

Patterns we hear on Seed Hunter fit calls:

• Size stated in dollars with no denominator ("put $40k on it" — of what?).
• Size stated as "percent of the trading account" while payday cash and rent reserve live elsewhere.
• Max loss described as a price percent while the instrument was convex or levered.
• Averaging down without rewriting the original budget percent.
• Copying tip-group size as if social volume were a risk committee.

None of those become ethical by being common. They become expensive by being undocumented.

Also related: when your max-loss exit fails—budget and exit are siblings. A percent of liquid NW without an exit clause is a dare.

A past-tense budget packet you can reuse

After an asymmetric experience—win, lose, or messy—log something like this:

1. Liquid net worth band at entry (round numbers are fine; vanity precision is not).
2. Dollars at risk and the implied percent of that liquid base.
3. Instrument and whether max loss was capped (defined risk) or open-ended.
4. What would force a rewrite of the budget midstream (financing call, unlock, dilution).
5. Outcome vs the budget story you told yourself.
6. Would-repeat rule in one sentence that a skeptical peer could check.

If that packet feels embarrassing to write, that is usually the useful part.

Liquid NW vs "I can always raise"

A second failure mode sits beside opaque leverage: recourse fantasy. People write budgets as if friends, future bonuses, or a HELOC were part of liquid NW. Those may be real resources later. They are not liquid NW at entry unless they are already dollars you can lose without a negotiation.

When Archegos asked brokers for a standstill while liquidating, according to the Credit Suisse investigation summary, the standstill did not hold and default notices followed. Retail versions are quieter: hoping a payroll lands before a margin call; hoping a spouse will "understand." Log hopes as hopes. Do not put them in the denominator.

Comparing experiences without vanity precision

You do not need audited financial statements to use percentages. Bands work: "about 8–10% of liquid NW" is more honest than "exactly $47,250 of a $512,100 household." Vanity precision usually arrives when someone is performing diligence rather than doing it.

Across a year of ConvexClub notes, percent-of-liquid-NW makes unlike domains comparable. A real-estate deposit, a concentrated equity ticket, and a crypto unlock position stop pretending they are different species of risk just because the websites look different. The journal across domains idea is the same muscle.

What we are not saying

We are not saying Credit Suisse's $5.5bn loss proves any percent band for you. We are not saying family offices are a model. We are not saying total-return swaps belong in a retail product pitch. The public case is a loud reminder that size without a durable denominator is a process bug—whether the notional is twelve figures or five.

If you need a rule of thumb, invent your own and write it down before entry. Then grade yourself against your rule, not against someone else's tweeted risk paradigm.

Worked reconstruction (labeled composite)

Composite: M puts $25k into a concentrated public-markets thesis. Liquid NW roughly $200k. They never write "12.5%." They write "small account, high conviction." Six weeks later the mark is −35% and they average in another $15k because "the thesis is more obvious." The second ticket was a silent budget rewrite to ~20% without a new packet.

The lesson is not "never average." The lesson is: averaging is a new experience that needs a new percent, a new shape check, and a new would-repeat rule. Undocumented averaging is how conviction launders overrun.

Try this yourself

Pick one past concentrated bet. Write its budget as a percent of liquid net worth at the time, not as a mood. Log the reconstructed packet as a past-tense experience on ConvexClub—empty fields count as findings. If you want partner context later, see Partners.

Next steps

Start free · Pro · Five factors