Lucius Quinctius Cincinnatus

28 | Financial Analyst | Investor On the Road to $1 million

[Lucius Quinctius Cincinnatus @LukasTuborg]

Idealised math before fees, financing, cash yield and tracking:

Flat round trip: $SSO gains 20%, then loses 18.18%, ending at 98.18. A half-weight barbell ends at 99.09.

Smooth rally: $SSO gains 20% twice, ending at 144. A half-weight barbell ends at 122 versus 121 for the index.

ProShares targets 2x the S&P 500’s daily return and warns that multi-day results depend on compounding and volatility.

[Lucius Quinctius Cincinnatus @LukasTuborg]

50% $SSO plus 50% cash is not a synthetic S&P 500 fund.

It starts with the same 1x daily exposure. But after +10% then -9.09%, the index is flat while the barbell loses 0.91%. After +10% twice, the index gains 21%; the barbell gains 22%.

Same starting beta, different path.

Leverage can preserve liquidity, but daily resetting makes volatility part of the allocation.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Assumes the same DKK 600k real replacement cost in five years, a 4% real hurdle rate, and no interim failure, maintenance, insurance or energy-cost difference.

DKK 600k / 1.04⁵ = approximately DKK 493k.

The DKK 107k is option value, not guaranteed savings. Failure risk, construction-price inflation and energy losses can justify earlier replacement.

[Lucius Quinctius Cincinnatus @LukasTuborg]

The cheapest time to add daylight may be during a reroof. That does not mean the cheapest time to reroof is now.

At a 4% real hurdle rate, DKK 600k spent five years from now has a present value of DKK 493k. A serviceable roof carries DKK 107k of deferral value.

Bundle private upgrades when replacement is due. Do not accelerate the base project just to access the bundle.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Assumes 25% Danish VAT, company-owned inventory sold to Danish consumers, and no channel costs.

175 / 1.25 = 140
150 / 1.25 = 120; 140 / 120 = 1.167
100 / 1.25 = 80; 140 / 80 = 1.75

SKAT states that Danish VAT is generally 25%. Company sales belong in the VAT accounts, while transferring inventory into private use can itself trigger withdrawal VAT.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Clearance pricing should be compared on expected cash, not sticker price.

A DKK 175 webshop sale leaves DKK 140 before fees after Danish VAT. A DKK 150 marketplace sale leaves DKK 120, so it only wins if the chance of selling is at least 17% higher. At DKK 100, it needs to be 75% higher.

A marketplace discount is rational when distribution converts enough dead inventory into cash.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Assumes 90% equities, 10% non-yielding cash, no tax or trading costs.

Down: 90 × 0.80 + 10 = 82. Recovery requires +25%, giving 102.5.

Up: 90 × 1.20 + 10 = 118 versus 120 fully invested.

Cash yield reduces the cost. Poor deployment reduces the payoff.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Cash is not diversification. It is a paid timing option.

Hold 10% cash. If equities fall 20%, deploy it, then the market recovers: the portfolio finishes at 102.5 versus 100 if fully invested.

If equities rise 20% first, the cash portfolio reaches 118 versus 120.

You paid 2 points of upside for 2.5 points of drawdown optionality.

[James E. Thorne @DrJStrategy]

How soon we forget 2018.

Fed Incompetence in a Midterm Year? What a Surprise.

Jerome Powell’s 2018 mistake was not simply a bad forecast. It was projecting confidence he could not justify and leaving investors to absorb the consequences.

During the midterm-election year of Donald Trump’s first presidency, Powell discussed the trade war’s inflation risks. Yes, he viewed tariffs as causing inflation.

Then, on October 3, interviewed by PBS’s Judy Woodruff, he declared interest rates were “a long way from neutral” and might have to go beyond it.

Inflation was near target. Tariff effects were not yet evident. Powell nevertheless turned an uncertain estimate into a threat of substantially tighter money, weeks before Americans voted.

Bond yields jumped. Equities were hammered. The S&P 500 lost 13.5 % in the fourth quarter. After the 2018 midterms, long-term Treasury yields fell dramatically as the tightening scare became a growth scare.

Powell compounded his failure. December brought another increase and his description of quantitative tightening as running on “automatic pilot.” Investors faced deteriorating conditions; their central banker defended the machinery.

January finally brought patience. Powell discovered discretion only after markets punished his rigidity.

Now Kevin Warsh’s Fed has raised rates before another midterm election. Inflation is higher, but toughness still does not equal judgment.

The long end of the yield curve is screaming higher. That is no endorsement of central-bank competence. A Fed that tightens short rates without reassuring long-term lenders deserves scrutiny, not applause.

Today’s saving grace is the investment engine. Trump enacted supply-side incentives early, improving the economics of equipment, research and manufacturing investment.

Yet, an oil price shock, and tariffs have the Fed hiking rates into a negative growth shock, while trimmed mean PCE is at the 2% range. And yes they, the Fed want more hikes!!

A generational AI cycle is powering extraordinary earnings growth.

Analysts expect S&P 500 profits to rise 35% this year.

Provides the stock market enough fundamental support to counter act an incompetent Fed.

That cushion does not excuse another monetary blunder.

Expanding productive capacity is a source of strength, not a problem for the Fed to crush.

Amazing how Fed incompetence keeps rearing its ugly head in midterm election years.

[Andreas Steno Larsen @AndreasSteno]

https://t.co/blj8YAWnHk

[Lucius Quinctius Cincinnatus @LukasTuborg]

Example: assume an as-is asset is worth DKK 43k/m² today and a genuinely comparable finished asset is worth DKK 58k/m².

At exit, apply the same market assumption to both. The resulting spread belongs to the renovation thesis. General market appreciation does not.

Added area should enter only after ownership, permitted use and legal registration are sufficiently certain. This is a valuation framework, not an appraisal.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Renovation returns are often overstated by mixing market appreciation with value creation.

Model two future values:

Today’s as-is value × market CAGR
Today’s finished comparable × the same CAGR

The difference at exit is the renovation spread. Then subtract capex, financing and execution risk.

If the project adds legal m² or changes the asset class, build separate bridges for those too.

Otherwise ordinary market beta gets reported as construction alpha.

[Lucius Quinctius Cincinnatus @LukasTuborg]

The OpenAI figures are reported, unverified and not directly comparable: Anthropic includes cloud-partner sales that OpenAI excludes.

Micron separately disclosed $150bn of RPO based on committed volumes and minimum pricing, 26 take-or-pay agreements, 35%+ estimated revenue coverage through 2030 and $12.7bn of customer deposits.

Inference: contractualised memory revenue may be less sensitive to one lab’s run-rate metric. Micron does not identify the customers, and Samsung’s preliminary results were a concurrent pressure on memory shares.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Yesterday’s AI selloff treated a metric-definition discrepancy like a hardware-demand cancellation.

OpenAI’s annualised revenue was reported near $50bn, not the $70bn media estimate built partly to match Anthropic’s cloud-inclusive metric. $MU fell 4.8%, more than $NVDA’s 2.9%.

Yet Micron has $150bn of minimum-price RPO, 26 take-or-pay agreements covering 35%+ of revenue through 2030 and $12.7bn of customer deposits.

Memory was repriced without any disclosed contracted-customer cancellation.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Illustration assumes DKK 4.2m principal, 2.8% annual interest plus contribution, monthly payments and a fresh 30-year amortising term.
Payment on DKK 4.2m: DKK 17,258/month
Interest-only cost: DKK 9,800/month
Principal supported by DKK 20,000/month: DKK 4.867m
Incremental capacity: DKK 667k
Excludes refinancing costs, tax effects, changing contribution rates, affordability tests and loan-to-value constraints. This is a financing model, not a quote.

[Lucius Quinctius Cincinnatus @LukasTuborg]

A mortgage extension does not lower the economic cost. It monetises future amortisation.
At 2.8%, resetting DKK 4.2m to 30 years gives a DKK 17.3k monthly payment. Keep paying DKK 20k and the same cash flow supports DKK 4.87m, releasing about DKK 670k for renovation.
The capacity is borrowed from the future repayment schedule.
That can work, but only if the capex creates more value than the principal you stopped repaying.

[Department of State @StateDept]

HAPPENING NOW: Secretary Rubio delivers remarks from the Acropolis in Athens, Greece https://t.co/XHBFbU4n7S

[Lucius Quinctius Cincinnatus @LukasTuborg]

SparEnergi’s examples use 30 m² of one-leaf windows with old two-layer thermal glazing. Pane-only replacement saves 3,000 kWh/year; full A-rated windows save 4,200 kWh/year.

At DKK 0.93/kWh, the incremental 1,200 kWh is worth DKK 1,116/year.

Inference: pane-only may dominate when frames are structurally sound. Full replacement can still win on rot, leakage, noise, maintenance or aesthetics.

[Lucius Quinctius Cincinnatus @LukasTuborg]

Replacing a sound window frame can be an expensive way to buy the last 29% of the modelled energy savings.

SparEnergi’s own example: across 30 m² of old double glazing, pane-only replacement saves 3,000 kWh/year. Full A-rated windows save 4,200 kWh.

The new frames add just 1,200 kWh/year, about DKK 1,100 at Denmark’s current average district-heating price.

If the timber is sound, underwrite glass and frames separately.

[𝐀𝐠𝐫𝐢𝐩𝐩𝐚 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭𝐬 @Agrippa_Inv]

“and Mackenzie”… do you guys have no shame? That site isn’t even live yet.

And as Kent Draper clarified in response to your recent hit piece, their CUSTOMERS preferred faster access to compute and were willing to compromise on redundancy as $IREN upgraded the Prince George site from mining to AI cloud.

$IREN didn’t “cut corners”… customers wanted quicker access to compute and were willing to eat some temporary downtime in exchange as the site was being upgraded.

That’s why companies like TogetherAi, which you referenced in your original hit piece as your primary source, RENEWED / extended their contract with $IREN… oh the irony.

This is so misleading. Very convenient for you to leave this context out.

Honestly, I think it’s time for you to get sued for libel. Truly the TMZ of the data center industry... extremely pathetic.