Cash Is Air, Allocation Is Survival: Retirement Withdrawal Tiers

In 30 seconds: Chapter 3 in one line — cash is air, allocation is survival. The point of allocation isn’t an extra 1–2%; it’s staying alive in the worst case (a 70% crash and a family emergency in the same week) without being forced to sell at the bottom. Workers keep 6–12 months of cash and put the rest in QQQ; retirees pick one of four tiers by withdrawal rate.¹

Cash is air

You don’t get rich on cash, but you can’t live without it. James’s real example: a reader’s father suddenly had a stroke needing a ~$30k out-of-pocket stent — if you hold no cash and the market happens to be falling, what do you do?¹

• Still working: keep 6–12 months of living expenses in cash, the rest 100% in QQQ. That cash is your oxygen when unemployment and a crash hit at once.
• Retired: cash position 20–30%, depending on withdrawal rate (below).

📌 2026-06 update: cash flow ≠ cash. James later refined the point here from “how much cash to hold” to “can you keep generating cash flow” — retirement needs 15 years of cash flow (from dividends, selling money-market funds, or pledging), not 15 years of idle cash. See 15 years of cash flow.

“Cash is inefficient” is true — but do you want to pass through calmly, or accelerate while risking the whole boat sinking? Allocation is the garnish; keeping enough cash is the main course.

Lump sum or dollar-cost average? Buy at market, in one click

Once your ratios are set, buy it all at once, at market — don’t average in. Averaging leaves money on the sidelines waiting for a “perfect moment” that statistically doesn’t exist; bull markets are ~70% of the time, so waiting costs more than the price difference you think you’re saving. The only exception: you haven’t settled your ratios yet — then stop and think first.¹

“99% of people put way too much weight on how they buy in.” (Post 0039, 2026-07) How you buy in isn’t what decides success — holding for the long run and never selling is.²

A real case: in 2022 someone sold a house and asked how to deploy the money. James told them to average in — not because averaging is better, but because for someone not yet psychologically ready, going all-in at once is too strong a move; averaging only feels less scary about the market (and it’s only a feeling).²

But statistically, lump sum beats averaging. A large body of research shows lump-sum has a higher expected value than DCA, and the reason is probability — over the long run the market always rises:²

• Over the past twelve years (2009–2026) it rose almost every year; any time you invested a lump sum was right. The only top was mid-2021 into H1 2022 — a short-term loss — but averaging in didn’t beat lump sum there either: spread over twelve months, you’d have bought right up the peak the whole way.
• A rough probability estimate: across ~17 years (2009–2026) only half a year (H1 2022) fell — under a 3% chance of a decline, so a ~97% lump-sum win rate; even counting from the 2000 peak, with 3.5 down years in 27, the win rate is still ~87%. (James’s rough framing, not a rigorous backtest.)

The only scenario where averaging makes sense: you can foresee in advance a decline lasting more than a year — but in an index that rises perpetually, such windows are both rare and unforeseeable. Whether the market is topping, rising, or ranging, lump sum’s expected value beats averaging.²

And averaging is just as hard as lump sum. Market up, you want to wait for a pullback; market down, you want to see if it falls further — on average, averaging still doesn’t beat lump sum. The short run can’t tell you which buying method is better; only the long run reveals the power of the lump sum. If you truly can’t do lump sum and can only average, that’s fine — just don’t drag it out.²

There is no perfect way to buy in. However you buy, as long as you buy the US broad-market index whenever you have money and then hold for the long run and never sell, it will make you rich. Buying isn’t the decisive factor in investing success — sitting still like a mountain is. When you’re most afraid, be brave (echoing the crash is your best friend).²

Beginners: start with a conservative allocation (fewer than 100 videos watched)

“A worker keeps half a year of cash and puts the rest all into QQQ” is for someone who already understands and can hold on. Beginners are different — 00552 gives a more conservative starter allocation:³

• When you’ve watched only a few videos, you’re still “a blind man feeling an elephant”; it usually takes about 300 videos before you even roughly understand investing. At this stage, don’t be overconfident, and don’t go aggressive too early — “cash is air; without air, everything else is fake.”
• Beginners who’ve watched fewer than 100 videos — regardless of prior investing experience — should start from one of the three below (the extra cash is so you can hold on before you’ve built conviction):

Starter tier	Nasdaq-100 index fund	Cash
70 / 30	70%	30%
60 / 40	60%	40%
50 / 50	50%	50%

🔗 This matches the seven levels: a beginner’s biggest risk is “thinking you get it and going aggressive too early” — one big drop washes you out. First survive on a high cash position, watch enough videos, build conviction, then gradually lower cash and raise Beta.

How much chili: allocation is personal

Allocation is like how much chili to add to noodles — age, family burden, risk tolerance, and tax regime all differ, so there’s no standard answer. James just gives you the “spice level” of each ingredient so you can mix your own:¹

• QQQ growth: ~12–15%/yr
• QQQI dividend yield: ~1%/month
• Cash return: the central bank overnight rate (≈ short-term bills / money funds)

Retirement withdrawal tiers (four)

How much you withdraw in retirement decides your tier:¹

Tier	Withdrawal	Fits assets of	Allocation
1	2%/mo	≈ 50× annual spend	80% QQQ + 20% bills/MMF (no pledge)
2	3%/mo	≈ 33× annual spend	70% QQQ + 30% bills/MMF (no pledge)
3	2–3%	live on [[質押借款	pledged loans]]
4	> 3%	retiring under 33×	100× monthly spend in QQQI + rest 70/30

• Tier 3’s 3% pledged version holds less QQQ (65%) than the no-pledge version (70%) — because pledging itself adds leverage, so the equity sleeve yields some room to the safety cushion.
• Tier 4 uses QQQI (monthly distributions from dividends + covered calls) for cash flow; the downside is lower long-term return than pure QQQ, so it’s only needed above a 3% withdrawal rate. The remaining 70/30 then isn’t rebalanced — it’s an inheritance accelerator.

Where to park cash? US: BOXX / SGOV / BIL / MMFs (VMFXX, SPRXX); Taiwan: 00865B; China: 511880 (money-market ETF, ~1%) — don’t use 161115 as cash; its bond/stock holdings can lose principal, so it’s unsuitable as a cash position (James’s correction in 00567). Never put it in long-term or corporate bonds, and never swap cash for high-dividend funds (see 年輕人適合買高股息嗎) — cash’s job is a “liquidity parking lot”: don’t lose value, stay available.¹

🧊 Why short bonds, not long bonds? (00559 / 00564) This is the key to “a short bond equals cash”:⁴

• Short bonds (returning principal in 1, 3, or 6 months) ≈ cash: within three months the odds of a US default or a big inflation swing are near zero, the risk-premium moves are tiny, and you’re guaranteed your principal back.
• Long bonds (returning principal only in 5, 10, 20, or 30 years) = risky assets: lend the US 30 years and you bear 30 years of inflation and default risk; a small market wobble and your 30-year Treasury can drop 20%+ (which is also why 30-year mortgage rates exceed 15/20-year ones).
• So “hold cash” means short bonds / money market, not long bonds standing in for cash; long bonds swing like stocks and provide no oxygen. For why long bonds and corporate-bond funds shouldn’t count as a safe asset at all, see 為什麼不買債券.

The closing line: simple is the correct solution. No leveraged funds, no “smart” rebalancing — if you pledge, just rebalance mechanically.

⚠️ Summarized from CLEC’s teaching for education only — not personalized advice. Ratios, pledging, and withdrawal rates depend on your situation and local tax law; retirement scenarios must be paired with the three lines of defense.

Footnotes

1. CLEC James, The Ten-Billion-Dollar Investment Lecture, Chapter 3, May 2026. Converted source: raw/docs/教學資料/價值十億元的投資講座_無圖版_v1.pdf. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

2. CLEC James, Post 0039 “99% of people put way too much weight on how they buy in,” 2026-07-23, X @CLEC168 long-form. The win-rate figures are James’s rough framing in the post (down years ÷ total years), not a rigorous backtest; the core takeaway is “how you buy in doesn’t matter, holding long-term does.” Pasted in manually; not yet in raw/docs/X及YouTube的貼文/. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

3. CLEC James, 長篇 00552 “The deadly risk you don’t know about… taxation, inheritance, and wrong allocation,” 2026-02-07, beginner starter allocations (70/30, 60/40, 50/50 for those with fewer than 100 videos watched); ~300 videos to roughly understand investing. Matching deck page in raw/docs/簡報資料/00552…. Transcript in raw/transcripts/長篇/00552…. ↩

4. CLEC James, long session 00559 “Dollar Hegemony… Retirement Cash Flow,” 2026-04-04, short-vs-long bond explanation around @28:30; long session 00564 “The Biggest Risk Isn’t Volatility…,” 2026-05-09, same deck slide. Transcripts at raw/transcripts/長篇/00559…, raw/transcripts/長篇/00564…. ↩