Sharpmarkprep

Buy-side interviews don't give second chances.

Make your prep count. Practice tests for HF / PE / IB recruiting, from real experience.

Start today for free.

[Sharpmarkprep @sharpmarkprep]

@TheProfInvestor Markets are extremely frothy right now, conviction trades will win

[Sharpmarkprep @sharpmarkprep]

@DrJStrategy Sheep to the slaughter

[Sharpmarkprep @sharpmarkprep]

Buy-side final rounds are quietly moving back offline. No internet, no phone, model on a laptop in a room.

Not because firms are anti-AI. Most of them expect you to use it on the job. It's because a take-home stopped telling them anything. Everyone's output looks competent now, so output stopped being the signal.

What's left is whether you can defend it. Where did that margin assumption come from, why that exit multiple, what breaks if you're wrong.

AI made the work easy to produce and the understanding easy to fake. So they stopped testing the work.

Free timed practice test if you want reps under those conditions. Link's in the bio.

[Sharpmarkprep @sharpmarkprep]

Over 1.2m South Koreans just got margin called this week. That's 1 in 30 working age adults.

Meanwhile the S&P is 2% off its high and people are calling it a crash.

Leverage decides how much a selloff hurts. Not the size of the move.

[Sharpmarkprep @sharpmarkprep]

@InTheAssembly Need an inverse cramer etf asap

[Sharpmarkprep @sharpmarkprep]

@Kalshi_Finance History's way of saying 'good luck with that DCF.'

[Sharpmarkprep @sharpmarkprep]

@jun_song Every quarter: 'AI will replace finance roles.' Also every quarter: my model still has a circular reference I can't find. We're fine, everyone.

[Sharpmarkprep @sharpmarkprep]

@Alibaba_Qwen 2.4T parameters and it's still not sizing my WACC assumptions correctly. Job's safe for now... phew

[Sharpmarkprep @sharpmarkprep]

Semiconductors dragging the whole market down again. AI hype hitting some serious doubt on the capex side. Nasdaq getting wrecked while everything else hangs on.

Is this rotation or a reality check?

$NVDA $SOX $MU

[Sharpmarkprep @sharpmarkprep]

Something most people don't realise: the average hedge fund analyst isn't right more often than you. They're right about the same amount.

The difference is what happens on the wrong ones. Position sizing, correlation limits, and a hard exit rule mean a bad call costs a fraction of what a good one makes. That's the entire edge. Not better prediction but better survival.

It's why the industry survives decades whereas the market chops sideways and retail gets shredded.

Being right is common. Being right and still solvent is the actual skill, and it's trainable.

[Sharpmarkprep @sharpmarkprep]

@TradexWhisperer Typical sell-side pulling forecast to the right. Every quarter higher than the last is the part that would worry me, not excite me.

[Sharpmarkprep @sharpmarkprep]

@CKCapitalxx Buy the crash (crash not included)

[Sharpmarkprep @sharpmarkprep]

@awealthofcs This is what a correction looks like from the inside: everyone whose whole book was the same five tickers thinks the world's ending while the index shrugs.

[Sharpmarkprep @sharpmarkprep]

Love that we've collectively decided circular financing is a moat now. Sell the rod, fund the fisherman, buy the fish back, book it three times, and all accretive apparently. Incredible businesses...

Until the music stops.

[Sharpmarkprep @sharpmarkprep]

@blueprintsmb22 The gap between "can pitch" and "can print" is wider than anyone coming from banking wants to believe. Two totally separate skill trees.

[Sharpmarkprep @sharpmarkprep]

So TSMC dropped their earnings this week and it was genuinely great quarter. Beat on revenue, raised guidance, bumped full-year growth above 40%. On paper, everything you'd want.

Yet the stock sold off.

Whole chip sector down ~4%.

And this trips up so many smart people. They think the game is "find the good company." It's not.

It's "find where the crowd is wrong." TSMC being good was already in consensus. The surprise was the $60-64B capex bill landing on a market already backing out of AI.

Being right about the business isn't the edge. Being right about what everyone already believes is.

[Sharpmarkprep @sharpmarkprep]

@asklivermore The concentration angle is the quiet subplot here, money comes back to tech, but to fewer and fewer names each cycle. Curious how that thread keeps developing.

[Sharpmarkprep @sharpmarkprep]

Most people think hedge funds win by finding good companies. Good companies are the easiest thing on earth to find. Everyone already owns them, and everything good about them is already in the price. There's no money in being right about something the whole market already agrees on.

The money is in taking contrarian views to consensus and being right. When everyone's leaned the same formula, the same names, the same "obvious" trade, the setup for a violent move in the opposite direction is already built. The crowded side feels safe precisely because it's crowded. That's the trap. The pain trade is almost always the one nobody's positioned for.

Reading that is a real skill. Not "is this a good business," but "what does the market already believe, and where is that belief lazy or overcrowded." It's why buy-side interviews hand you a stock and ask for a pitch instead of a summary. They're not checking whether you know the bull case everyone knows. They're checking whether you can find the variant view and defend it when they push.

Most candidates never train that muscle. They walk in able to describe a company and walk out wondering why that wasn't enough. It was never the test.

That skill is trainable, and it's most of the game. I build the practice that gets you there. Link in the bio, first one's free.

[Sharpmarkprep @sharpmarkprep]

@awealthofcs Every name on that list was a consensus long at some point in the last two years. Feels less like a drawdown list and more a crowded-trade unwind list.

[Sharpmarkprep @sharpmarkprep]

@TW_trades_ Good results and a falling stock aren't really a contradiction. Stocks move on the gap between earnings vs. consensus, and at current multiples, consensus is so high that great isn't quite enough.