Premium Is Not Income: Four Ledgers Before Any Strategy

Published 2026-11-12 1998 words10 min read

Site Overview
Table of Contents
No table of contents
"Successful investors are often those who can remain calm when the market panics. —— Warren Buffett"

Premium Is Not Income: Four Ledgers Before Any Strategy

Cash Flow System · BlueprintNOK PMCC: two sides of one trade
$1,638.50

35 short calls, gross premium

LEAP unrealized
−$3,832.50
Rent coverage
43%
Account ledger
NAV − deposits

Last month my account collected more than $1,600 in premium.

At the same time, the account carried more than $3,800 of unrealized losses.

Both numbers are true. Only one of them can be called income.

This is not wordplay. Ninety percent of the self-deception in a cash-flow system happens on the road between “the premium landed” and “this trade made money” — the part nobody wants to look at. This is the blueprint for the column: fix the four ledgers first, then talk about strategy.

1. My first mistake: premium ÷ capital

My earliest math went like this: $380 of premium this month, $5,000 of capital tied up, a 7.6% monthly return, 91% annualized.

I was thrilled. The problem sat in three places, and every one of them made the number bigger:

  1. The denominator counted “the money I thought was tied up,” not the margin the account actually had locked;
  2. The numerator treated cash received as profit earned, ignoring the unrealized loss on the other leg at the same moment;
  3. The time denominator assumed the trade repeated 365 days a year, as if every month collected the same premium and no trade ever lost.

Stack those three amplifiers and a system that actually earns low-teens annualized can print a three-digit number.

The nastier part: this mistake does not lose money immediately. It just keeps you on the wrong road longer — because every month the “income” line grows while only NAV shrinks.

2. One real ledger: $1,638 of premium, $3,832 of unrealized loss

Take my NOK position. It is the most typical PMCC in the account:

  • 35 LEAPs as the base, 35 short calls on top;
  • On 10/20 I sold 25 of the 12C at $0.503, collecting $1,257.50;
  • On 10/26 I sold 10 more 12C at $0.381, collecting $381.00;
  • Gross premium: $1,638.50.

Over the same period those 35 LEAPs were down $3,832.50 (−49%).

Rent coverage = 1,638.50 ÷ 3,832.50 = 43%.

In other words: the “rent” I collected on NOK covered only a bit more than four-tenths of the loss caused by the underlying moving the wrong way. The rent is real. The unrealized loss is real. But the sentence “I made money by collecting rent” is false at the account level.

In an earlier piece I put the conclusion this way: the rent does not cover the house (full ledger). This piece takes one more step — the question is not whether to collect rent, but which ledger you use to record it.

Silent’s Note: None of the four ledgers below is “wrong.” What is wrong is using a number from ledger A to answer a question only ledger B can answer.

3. Four ledgers, four different questions

LedgerFormulaQuestion it answersHow it deceives you
A. Cashnet premium ÷ margin used ÷ days held × 365Is this capital working efficiently?Inflated by high IV; mistaken for a return
B. Account(NAV − net deposits) ÷ net depositsDid I actually make money this year?Noisy short-term; meaningless for one month
C. Realizedcumulative closed P&L (buybacks and commissions in)How much actually landed?Ignores risk still open in the book
D. Coveragenet premium ÷ unrealized loss on the same underlyingCan rent rescue a wrong direction?Only valid within one underlying, never across

Ledger A — cash: is this capital working efficiently?

This is the only ledger suitable for cross-comparison — comparing capital efficiency across underlyings and expiries in the same account.

A worked example with assumptions (everything below is hypothetical, not one of my fills):

  • Underlying at $20; sell a 21-day $18 put at $0.42 per share → $42 per contract;
  • Cash-secured margin: $1,800 (= 18 × 100);
  • Return on capital for the trade: 42 ÷ 1,800 = 2.33%;
  • If repeated seamlessly 17 times a year (365 ÷ 21), annualized: 39.7%.

That 39.7% only means something if three assumptions hold at once: the underlying does not move for 21 days, IV does not change, and every trade fills at this price. In reality those three almost never hold together — so I never treat it as expected return. I only use it to answer “is it worth parking $1,800 here?”

It deceives in a very concrete way: when IV goes from 30 to 60, the same delta and the same expiry nearly double the premium, and the ledger-A annualized number doubles with it. Your win rate did not double; your risk went up. Theta efficiency is essentially a shadow of IV (same delta, same DTE: theta per dollar depends only on IV).

Ledger B — account: did this year actually make money?

One formula: (NAV − net deposits) ÷ net deposits.

The numerator is the change in NAV; the denominator is the money you actually put in. It ignores premium, ignores “unrealized gains,” ignores annualized numbers. Its advantage is that it cannot be dressed up — you can collect a lot of premium in a month and lose money on the account ledger, and both can be true at once.

My account snapshot for 10/26 (read from TWS, synced 2026-10-27 02:24):

ItemValue
NAV$6,357.77
Cash$1,992.00 (31% of NAV)
Maintenance margin used49.4%
Leverage1.31x
Available margin$3,217.61

Compare 10/14: maintenance margin 75.2%, cash only 24% of NAV, delta exposure 2.14x, cash on hand $364. That day I sold 25 of the 12C for $1,255.80 and cash went from $364 to $1,620 — that trade was not a bearish bet on NOK; it was buying oxygen for the account.

The same trade is “rent efficiency” on ledger A and “pulling the account back from a forced liquidation to a decision I control” on ledger B. That is why the ledgers cannot be mixed.

Ledger C — realized: only what lands counts

Money is realized only after a buyback, expiry, or assignment and delivery.

The key is that buyback cost must be included. Sell at $0.50 and buy back at $0.10: net is $0.40, not $0.50; commissions come out too. Many people record only “premium received at sale” and put the buyback in a separate expense column — and then can never tell what the month actually earned.

My rule: net premium gets its own line and never enters performance statistics. Performance is ledger B only. The value of that rule is that it turns “collecting rent” and “making money” into two independent variables, so you can see which one is doing the work.

Ledger D — coverage: can rent rescue a wrong direction?

This ledger only works for one underlying: rent collected ÷ unrealized loss on that same underlying.

NOK is 43%. That means: if the underlying falls another similar amount, rent covers a bit more than four-tenths. So “collecting rent” is never a substitute for a directional view; it is a subsidy paid to the position.

Conversely, if a position’s coverage stays above 100% for a long time, that usually does not mean great rent collection — it means the underlying barely moved. Then the thing to worry about is whether you sold the rent too cheaply.

4. Four layers of the system, one switch each

Ledgers are the books; the system is the machine. My cash-flow system has four layers, each with a single switch:

LayerWhat it controlsSwitchRelated reading
CapitalHow much goes in, how much cash staysCash never below 25% of NAVThe 2% rule
PositionMaximum per underlying and per directionUnrealized loss per name ≤ 20% of NAV1.5x leverage cap
StrategyWhich structure collects the rentThe structure must price its worst caseAll-weather allocation · Selling puts
ReviewWhen to stop, when to reconcileMonthly reconciliation; coverage < 50% → cutSeven questions before entry

The easiest layer to skip is review, because it produces no income. Without it, the first three drift: positions get added because they feel easy, cash gets thinner, strategies get fancier. The value of a system is not “a few more right trades”; it is “fewer structural mistakes.”

5. What these ledgers do not solve

Boundaries have to be written down, or this becomes another marketing piece:

  1. The right ledger cannot fix a wrong direction. Sellers earn small and carry the tail; that math is in the impossible triangle. The four ledgers show you what you are earning and losing. They do not make the directional call.
  2. Margin ledgers are not comparable across account sizes. A $6,000 account and a $60,000 account at the same 49% margin usage have completely different staying power — the first has room for one move, the second for three. Ledger A only means something inside one account.
  3. Annualized numbers inflate when IV is high. A 40% annualized computed at IV 60 and a 15% computed at IV 25 are not the same risk. Never put ledger-A numbers from different IV regimes in one table.
  4. One month annualized is not a forecast. Compounding is fragile: the median wins a little, the tail loses a lot (ten-year projection). A good month times twelve is arithmetic, not expectation.
  5. 43% coverage is not “failure.” It only says this PMCC’s rent cannot cover its directional loss. Whether to keep holding is a separate question (three switches).

6. The five-minute monthly reconciliation

  1. Export two numbers from the broker: realized P&L and cash balance;
  2. Compute NAV − net deposits and record it as the month’s real performance (ledger B);
  3. Record net premium on its own line (ledger C, after buyback cost and commissions);
  4. For each position, compute coverage = net premium ÷ unrealized loss on the same underlying (ledger D);
  5. Write one row: month / NAV / net deposits / realized / net premium / lowest coverage / largest single-trade risk.

Keep that table for six months and you will see an account that looks nothing like your memory — memory is full of “look how much rent I collected,” while the table shows whether NAV actually moved.

7. One line to take away

Premium is a premium, not profit; NAV minus net deposits is the only report card that counts.

Get that number right before picking strategies. Reverse the order and every extra hour of effort takes you further off course.


Data & Sources

  • Updated 2026-11-12 | First published 2026-11-12
  • Account and positions: TWS snapshot read directly (synced 2026-10-27 02:24 Singapore time), not manually entered
  • Fills and premium: IBKR execution reports (four NOK 12C fills on 2026-10-26, 13:29 / 14:03 ET)
  • Option IV and Greeks: implied from fill prices via Black-Scholes (r = 4%, no dividends), cross-checked against TWS modelGreeks
  • The 39.7% example in section 3 is purely hypothetical ($20 underlying, 21 DTE, $0.42 premium, $1,800 margin, repeated seamlessly 17 times), not any real fill

Risk Boundaries

This piece covers accounting ledgers and system structure only. It is not investment advice and promises no return. Option sellers can lose far more than the premium collected in extreme conditions: earnings gaps, short squeezes, and further declines after assignment can all turn “rent” into a very small number. No ledger replaces an estimate of the worst case — price the maximum loss first, then discuss how much rent you can collect.

Related posts