← Home  ·  A Starter Guide

Investing Basics

A plain-English walkthrough of how to think about stocks, build a portfolio, and avoid the most common mistakes new investors make.

01The right mindset

Investing is not gambling and it's not get-rich-quick. You are buying small pieces of real businesses. Over long periods (10+ years), the stock market has averaged roughly 7–10% per year after inflation. That's the prize for being patient.

The single biggest edge a small investor has over a Wall Street pro is time. You don't have to perform every quarter. You can sit on a good company for a decade. Use that advantage.

Core principle

Think in years, not days. Most of the noise on financial news doesn't matter to a long-term investor. Tune it out.

02Building a portfolio

A good portfolio is diversified — spread across enough things that no single bad bet can wipe you out, but concentrated enough that your winners actually move the needle.

ETFs vs. individual stocks

You want a mix of both, but the right ratio depends on how much time you'll actually spend researching. If you don't want to read 10-Ks, lean ETF-heavy. If you do, individual stocks are where the real outperformance lives — that's the path I take.

ETFs (the floor)

A basket of many stocks bundled into one ticker. Owning VOO or VTI means you own a slice of the entire market. Low fees, instantly diversified. They keep you in the game if your individual picks underperform.

Individual stocks (the engine)

Picking individual companies can dramatically outperform the index — but requires research and conviction. The real money is made by owning a few great businesses for a long time, not by spreading thin across hundreds.

How I allocate (an active investor's framework)

This is more concentrated than a textbook portfolio. The trade-off: more upside if you're right, more volatility along the way. If you don't want to spend time researching, flip it — go 70% ETFs and skip the speculative sleeve.

Position sizing

No new individual stock should be more than ~8–10% of your portfolio when you buy it. If a winner grows past that, great — let it. But don't start overconcentrated. Speculative bets should be small — 0.5% to 2% positions. You own a basket of them so even if half go to zero, the survivors pay for the rest.

Diversify across sectors, not just tickers

Don't own five tech stocks and call yourself diversified. Try to spread across sectors that move on different drivers — tech, healthcare, financials, energy, industrials, consumer. When one zigs, another zags. The portfolio I'll show you in the next sections is admittedly tech-heavy, because I believe in the AI buildout — that's a conscious bet, not an accident.

03Getting in — don't go all at once

The biggest mistake beginners make is dumping their entire cash pile into the market on day one. That's market timing in disguise — and you'll have no dry powder if the market drops 15% next month.

Dollar-Cost Averaging (DCA)

Instead of buying $10,000 of a stock today, buy $1,000 a month for 10 months. Some months you'll buy at higher prices, some lower — but on average, you smooth out the volatility and remove the emotional pressure of "is now the right time?"

Why DCA works

It enforces discipline. You buy mechanically regardless of headlines, which is almost always better than trying to guess tops and bottoms. The market is unpredictable short-term but trends up long-term.

Keep some cash on the side

Always keep a portion uninvested — call it 5–15% — so when the market panics and good companies go on sale, you have ammo to buy. Cash is a position too.

Add more on the dips

When a quality stock you already own drops 15–25% on broad market fear (not on a real problem with the business), that's often a chance to add to your position at a discount. This is the opposite of what most people do — they sell into fear.

04How to evaluate a stock

Before you buy a company, you should be able to answer four questions: What does this business do? Is it priced fairly? Is the momentum on its side? Is the sector growing?

1. Fundamentals — is this a real business?

2. Valuation ratios — is the price reasonable?

The three ratios that actually matter most of the time: P/E, P/S, and PEG. The catch — they only mean something when you read them alongside the growth rate. A "high" P/E is fine for a company growing 40% a year; a "low" P/E is a trap if growth is going negative.

P/E ratio Price ÷ trailing 12-month earnings. Lower = cheaper, but only relative to peers and growth. S&P 500 averages around 20×. Mature companies sit lower, growth names higher.
Forward P/E Same idea using next year's expected earnings. This is far more useful for fast-growing companies — a stock might look expensive on trailing P/E (e.g. 60×) but cheap on forward P/E (25×) because earnings are about to jump. Always check both, and lean on forward P/E for growth names.
P/S ratio Price ÷ sales. The right tool when a company isn't profitable yet but is growing fast. Compare against peers — software is normally 8–15×, industrials 1–3×.
PEG ratio P/E ÷ growth rate. Built specifically to compare expensive fast-growers against cheap slow-growers. Under 1.0 generally signals a reasonably-priced growth stock. Over 2.0 = the price is ahead of the fundamentals.

Always check the growth rate. Revenue growth tells you if the business is actually expanding. For any stock you're buying because of growth, ignore today's earnings number — focus on whether revenue is growing 20–40%+ a year and whether margins are improving. The market pays a premium for growth because growth compounds. A 30% grower at a 40× forward P/E is often a better buy than a 5% grower at 15×.

3. Trend & momentum — when to actually buy

Fundamentals tell you what to buy, trend and momentum help with when. A great company in a brutal downtrend can keep going down. You don't need to catch the exact bottom — but you also don't want to buy something running parabolic at the top. Two steps:

Step 1 — Check the trend with SMAs (Simple Moving Averages). An SMA is just the average price over the last N days. The two that matter:

Simplest rule for new investors

Only buy stocks trading above their 200-day SMA. You'll miss some early bottoms but you'll avoid the catastrophic mistake of catching falling knives — which is what wipes out most beginner portfolios.

Step 2 — Time the entry with RSI (Relative Strength Index) — a 0–100 score of how overbought or oversold a stock is over the past 14 days. Within a confirmed uptrend, this tells you when to add.

The combination is what makes this work: SMA tells you the direction, RSI tells you where in the move you are. A stock with RSI 30 above its 200-day is a buyable dip; the same RSI 30 below its 200-day is often a falling knife.

4. Sentiment — what is the market feeling?

Sentiment is the mood around a stock. When everyone is euphoric about a name, it's often near a top. When everyone hates it but the business is still solid, that's often opportunity. Warren Buffett's line: "Be fearful when others are greedy, and greedy when others are fearful."

5. Smart money signals — insiders & institutions

Two of the most underrated tools for retail investors: tracking what insiders (executives, directors) do with their own money, and what institutions (hedge funds, Berkshire, mutual funds) are buying and selling. These are public SEC filings — anyone can look, very few do.

Insider trading (the legal kind)

Every time a company insider buys or sells their own stock, they file Form 4 with the SEC within 2 days. You can see exactly who, when, how much, and at what price.

Institutional trading (13F filings)

Any fund managing over $100M has to disclose its holdings quarterly via a 13F filing. There's a 45-day lag, but you can see exactly what Berkshire, Pershing Square, Tiger Global, ARK, Citadel, and every other big fund owns.

Where to actually look — all free

OpenInsider — best for Form 4 insider buying/selling data, filterable by company or insider.
Whalewisdom or Hedgefollow — 13F holdings by fund or by stock.
Finviz — each stock page has an "Insider Trading" tab.
SEC EDGAR — the raw filings themselves if you want primary sources.

How to use this in practice: not as your primary signal — fundamentals and trend come first. But once you've decided you like a stock, check whether insiders are buying and institutions are accumulating. Both yes = conviction goes up. Insiders dumping and institutions exiting = conviction goes down, even if everything else looks fine.

6. Sector growth — is the tide rising?

A mediocre company in a booming sector often beats a great company in a dying one. Ask: where is this industry going over the next 5–10 years? AI, energy transition, aging demographics, automation — these are tailwinds. Print media, cable TV — headwinds.

05Themes I'm playing right now

This is the part where I get specific. Active investing isn't about picking 60 random stocks — it's about identifying 4–6 multi-year trends you believe in, then owning the companies best positioned to benefit. Below are the themes my portfolio is built around. None of this is a recommendation to buy these names — I'm showing you the thinking, not handing you a shopping list.

1. AI infrastructure — the picks-and-shovels bet

Everyone's racing to build AI, which means a massive buildout of data centers. Forget the AI apps — focus on what they physically need: chips, power, cooling, optical interconnect. The surest part of the bet.

Compute & chips: NVDA, TSM, ASML, AVGO, QCOM, AMKR
Memory: MU, SNDK
Power for data centers: CEG, VST, GEV, ETN, POWL
Cooling & racks: VRT
Optical interconnect: COHR, LITE, GLW, AAOI
Servers & cloud: SMCI, DELL, NBIS, IREN

2. Mega-cap quality tech — the compounders

The biggest tech companies have moats so wide they're essentially impossible to dislodge. Boring, but they're the engine that quietly does most of the work in a portfolio over 10 years.

NVDA, AAPL, MSFT, GOOGL, META, AMZN, NFLX, NOW, INTU, CRWD, PANW

3. Energy & critical minerals

The economy is electrifying — EVs, AI data centers, reshoring. That requires copper, nuclear power, and grid build-out. The "physical economy" side of the AI trade.

Nuclear & power: CEG, VST, GEV, SMR
Copper & minerals: FCX, COPJ (ETF)
Rare earths (speculative): USAR, UAMY, AREC

4. Crypto as a sleeve

A small allocation to Bitcoin and Ethereum as a non-correlated, scarce-asset hedge. Volatile, but the asymmetry over a 5–10 year window has been hard to ignore. Treat it like a position, not a religion.

BTC, ETH

5. High-conviction individual growth bets

Names with stronger conviction, sized bigger than average. Each has a specific thesis I can defend — defense AI software, a biotech with an approved drug, etc.

PLTR (AI/defense), TGTX (biotech), HIMS (telehealth), CELH (consumer), VIST (Argentine oil)

6. Frontier & speculative — lottery tickets

Tiny positions in moonshot ideas: quantum, satellite-to-phone, space launch, small modular nuclear, autonomous robotics. Most go nowhere. One or two might 10x. Each one small enough that being wrong doesn't hurt.

Quantum: IONQ, QTUM (ETF)
Space: ASTS, RKLB, PL
Nuclear (SMR): SMR
Drones & robots: ONDS, SERV
Lidar & sensors: OUST
Other: POET, BBAI, BE, NVTS
The thesis behind the themes

If you trace it back, almost everything I own is a bet on one of three things: (1) AI compute keeps growing → buy the picks-and-shovels, (2) electrification is the next decade's mega-trend → buy power, copper, nuclear, (3) a few specific companies have something special → concentrate there. Every position should map to a thesis. If it doesn't, you're collecting tickers, not investing.

06Quality vs. speculative tiers

Not every stock in a portfolio plays the same role. I think about my holdings in three tiers — core, growth, and speculative. The tier determines the position size, how I judge it, and what would make me sell.

Tier 1 · Core

Quality compounders — the bedrock

Target: ~50% of stocks · Position size: 3–10% each · Hold forever unless thesis breaks

Profitable, dominant businesses in growing industries. The kind of companies that survived COVID, the 2022 selloff, and every other scare. You don't trade these — you compound with them. Their job is to be steady so the rest of the portfolio can take risk.

BRK-B, NVDA, AAPL, MSFT, GOOGL, META, AMZN, NFLX, NOW, INTU, TSM, ASML, ETN, CMI, GLW, COHR, PANW, CRWD
Tier 2 · Growth

High-conviction growth — the engine

Target: ~35% of stocks · Position size: 1–5% each · Re-check thesis each quarter

Real businesses with real revenue, but more expensive valuations or higher execution risk. The story has to keep delivering. These are the names that drive outperformance if I'm right — and if I'm wrong, the damage is contained because each position is sized smaller. Earnings reports actually matter here.

PLTR, RDDT, TGTX, VST, CEG, GEV, POWL, VRT, NBIS, HIMS, CELH, SMCI, DELL, MU, SNDK, AVGO, QCOM, AMKR, LITE, ZETA, RKLB, IREN, VIST, FCX, IBM, BABA, APTV
Tier 3 · Speculative

Lottery tickets — asymmetric small bets

Target: ~15% of stocks · Position size: 0.1–1% each · Expect ~half to fail

Micro-caps, pre-revenue companies, frontier tech. Could each 5–10x or go to zero. The portfolio approach is what makes this work: spread across many, accept most will die, let the survivors carry the basket. Never size big — the rule is "if it goes to zero tomorrow, do I shrug or panic?" If you'd panic, it's too big.

IONQ, ASTS, PL, SMR, BBAI, ONDS, SERV, OUST, BE, POET, NVTS, AAOI, ALMU, AEHR, USAR, UAMY, AREC, SOFI, AMBA, VECO
A warning on Tier 3

Speculative bets are fun, which is dangerous. Don't let them creep above ~15–20% of the portfolio. And remember: the headline ones (quantum, space, fusion) are the ones with the most narrative — which means they're often the most overhyped. Read the actual financials.

07The watchlist

This is the heart of it — the full list of names I'm actually watching, grouped by what role they play. Screenshot the box below and you've got the whole watchlist on one screen. None of this is a recommendation to buy — it's the list I track and research, sized and timed using the rules in the sections above.

📋 The Watchlist

36 names · grouped by role

Foundation ETFs anchor it; staples do the compounding; sector picks and speculative names add the upside. Buy only above the 200-day SMA, on RSI pullbacks, DCA'd in over time.

Foundation · ETFs
QQQ
Nasdaq-100
ARTY
AI & Tech
QTUM
Quantum
DRAM
Memory / AI
NASA
Space economy
FLKR
S. Korea chips
Core staples · quality compounders
NVDA
Nvidia
GOOGL
Alphabet
MSFT
Microsoft
AAPL
Apple
META
Meta
BRK-B
Berkshire
PLTR
Palantir
RDDT
Reddit
Energy & power
CEG
Constellation
VST
Vistra
GEV
GE Vernova
Photonics & optical
COHR
Coherent
LITE
Lumentum
GLW
Corning
AAOI
Applied Opto
Semiconductors
TSM
TSMC
ASML
ASML
AVGO
Broadcom
Memory
MU
Micron
SNDK
Sandisk
AI infrastructure
NBIS
Nebius
VRT
Vertiv
SMCI
Supermicro
DELL
Dell
Software
NOW
ServiceNow
CRWD
CrowdStrike
PANW
Palo Alto
ZETA
Zeta Global
Defense
AVAV
AeroVironment
ONDS
Ondas
PLTR
Palantir
Space
ASTS
AST SpaceMobile
RKLB
Rocket Lab
PL
Planet Labs

Building the watchlist into a portfolio

The list above is what to watch. Below is one concrete way to turn it into actual position sizes — my real portfolio is more concentrated and a little riskier, this is a cleaner version with the same DNA. You don't need to own every name; pick 4–6 ETFs you like, 5–7 staples, and 2–3 sector picks where you actually have a view.

ETF foundation

~30% of portfolio

Diversified core + thematic exposure

6 ETFs

Your floor. Owns the index plus targeted AI, quantum, memory, space, and Korea-chip exposure.

QQQ
Invesco Nasdaq-100
Top 100 non-financial Nasdaq names. The core US tech-growth ETF.
ARTY
iShares Future AI & Tech
Diversified basket of AI beneficiaries across hardware and software.
QTUM
Defiance Quantum
Quantum computing and machine-learning theme.
DRAM
Memory / AI infrastructure
The chips that hold the data AI runs on. Memory has been one of the cleanest AI tailwinds.
NASA
Space economy
Satellites, launch, defense space — the commercial space buildout.
FLKR
Franklin FTSE South Korea
Samsung, SK Hynix. Memory and AI-chip exposure outside the US.

Staple stocks — quality core

~40% of portfolio

The businesses with the deepest moats and most reliable compounding. Hold them like real estate, not like trades.

NVDA
Nvidia
GOOGL
Alphabet
MSFT
Microsoft
AAPL
Apple
META
Meta Platforms
BRK-B
Berkshire Hathaway
PLTR
Palantir
RDDT
Reddit

Sector picks — 3 leaders per theme

~25% of portfolio

Pick the sectors where you actually have a view. Each of these is a high-quality leader in its space — not a speculative name.

Energy & power

3 picks

AI data centers and electrification need enormous power. These names sell that power or the equipment that generates and delivers it.

CEG
Constellation Energy
Largest US nuclear operator. Signing direct AI deals — including the Three Mile Island restart with Microsoft.
VST
Vistra
Utility with major nuclear + gas exposure. The biggest single beneficiary of AI-power demand.
GEV
GE Vernova
Gas turbines, grid equipment, wind. Builds the actual generation capacity.

Photonics & optical

4 picks

AI moves staggering amounts of data inside and between data centers. Light through fiber — not copper — is how that happens at scale.

COHR
Coherent
Optical transceivers and lasers. Key supplier for AI interconnect.
LITE
Lumentum
800G and 1.6T transceiver leader for data-center connectivity.
GLW
Corning
Optical fiber and specialty glass. Also tied to the data-center cabling buildout.
AAOI
Applied Optoelectronics
Laser and transceiver maker ramping data-center optics. Smaller and more volatile than the leaders above — the higher-beta way to play the same interconnect trend.

Semiconductors

3 picks

The chips themselves and the companies that make them physically possible.

TSM
Taiwan Semiconductor
The foundry that manufactures chips for NVDA, AAPL, AMD. Picks-and-shovels for the entire industry.
ASML
ASML
Only company on Earth that makes EUV lithography machines. Every leading-edge fab needs them.
AVGO
Broadcom
Custom AI silicon for hyperscalers plus networking chips. The "AI ASIC" winner.

Memory

2 picks

AI is insatiable for high-bandwidth memory (HBM) and storage. Every GPU server needs more of it — one of the cleanest AI tailwinds in the chip stack.

MU
Micron Technology
US memory leader. The HBM build-out for Nvidia GPUs is rewriting the cycle — Micron is one of only three HBM suppliers globally.
SNDK
Sandisk
Pure-play NAND flash storage, recently spun off from Western Digital. Direct beneficiary of AI storage demand and a tightening NAND supply cycle.

AI infrastructure

4 picks

Everything around the chip — racks, cooling, AI servers, and the cloud providers renting GPU compute to AI companies.

NBIS
Nebius
AI cloud / GPU compute, vertically integrated. Sells the actual compute that AI companies run on.
VRT
Vertiv
Liquid cooling and power infrastructure for AI racks. Cooling is the new bottleneck.
SMCI
Super Micro Computer
The highest-purity AI server vendor in public markets — custom-builds liquid-cooled GPU racks for hyperscalers. Volatile (had an accounting reset in 2024) but the underlying demand is real.
DELL
Dell Technologies
The bigger, more diversified AI server play. Steadier counterpart to SMCI with enterprise + PC exposure on top.

Software

4 picks

High-margin recurring revenue businesses that benefit from AI without having to fund the buildout themselves.

NOW
ServiceNow
Enterprise workflow automation + AI agents.
CRWD
CrowdStrike
Cybersecurity SaaS leader. AI-native detection.
PANW
Palo Alto Networks
Cybersecurity platform consolidating the whole stack.
ZETA
Zeta Global
AI-driven marketing and data cloud. Growing fast and recently profitable — a higher-risk, higher-growth name vs. the steadier three above. Size it smaller.

Defense

3 picks

Record defense budgets + the shift to drone & AI warfare are reshaping who wins. The legacy primes (LMT, RTX, NOC) are fine but slow — the modern drone & software names below are the better bets.

AVAV
AeroVironment
Drones and unmanned systems. How modern wars are actually being fought. Real revenue, real contracts.
ONDS
Ondas Holdings
Drone systems + private wireless networks for defense and industrial use. Smaller and more speculative — size it accordingly.
PLTR
Palantir
The defense AI software layer (Foundry, Gotham). The brain of modern military operations. Already in staples — counts as your defense exposure too.

Space

3 picks

Commercial space is finally turning into a real industry — direct-to-cell satellites, cheap launch, lunar landers. These are more speculative than other sectors here, so size them smaller (0.5–1.5% each).

ASTS
AST SpaceMobile
Direct-to-cellphone satellite network — your existing phone connects to space, no special hardware needed. Partnered with AT&T and Verizon. Pre-revenue but the constellation is being deployed.
RKLB
Rocket Lab
Electron rocket (second-most-launched US vehicle behind SpaceX) plus the Photon spacecraft platform. Building Neutron, a larger reusable rocket. Real revenue, real customers.
PL
Planet Labs
Operates the largest fleet of Earth-imaging satellites, selling daily geospatial data to governments, defense, and AI/analytics customers. Recurring-revenue model — a steadier way to own space than a pure launch play.
How to actually deploy this

Starting with say $10K: ~$3K spread across 3–4 ETFs, ~$4K across 5–7 staples, ~$2.5K across 6–10 sector picks, and ~$500 for any speculative names if you want spice. DCA it in over 3–6 months — don't dump it all in week one. Then add to it monthly with new savings, leaning toward whichever quality names happen to be on sale that month (RSI under 40, no broken thesis).

08When to sell (and when not to)

Let your winners ride

The most common mistake is selling winners too early. If you bought a great business and it's up 50%, the temptation is to "lock in the gain." But the biggest returns in investing come from holding multi-baggers for years. The math is brutal — selling a 10x winner at 2x means you missed 80% of the move.

Peter Lynch

"Selling your winners and holding your losers is like cutting the flowers and watering the weeds."

Trim, don't dump

When a winner gets euphoric — a parabolic move up, talking-heads everywhere, your barber giving stock tips — that's a good time to trim, not sell out. Take 20–30% off the table, lock in some gains, let the rest run. You stay in the game without being exposed to a violent correction.

When to actually sell

Don't sell because…

The market dropped 10%. There's a scary headline. A pundit on TV said it would crash. Your friend sold. None of these are reasons. They're noise.

09Common mistakes

10Quick checklist before you buy

For a Tier 1 or Tier 2 buy (core / high-conviction growth):

For a Tier 3 speculative bet, the bar is different:

11The bottom line

Buy good businesses. Buy them at fair prices. Spread your bets but don't over-diversify. Get in slowly. Hold your winners. Cut your losers when the thesis breaks. Ignore the noise. Repeat for decades.

That's it. The hard part isn't the strategy — it's the discipline to stick with it when the market is screaming at you to panic. Most people fail at this. The ones who don't end up wealthy.