I Know First
Weekly Newsletter | August 16th, 2026

• Read Here: The AI Portfolio Rebalances Wednesday: +45.37% vs S&P 500 +37.65% Since Inception
• Read Here: Leveraged ETFs Forecast Based on Deep Learning: Returns up to 7.85% in 3 Days

Need To Know First!

• ARK ETF Stocks Forecast Portfolio Based on a Self-Learning AI Algorithm: Returns up to 14.49% in 3 Days
• Top Technology Stocks Based on Genetic Algorithms: Returns up to 27.91% in 7 Days
• Quantitative Trading Based on Deep Learning: Returns up to 32.58% in 14 Days
• Tech Stocks Based on a Self-Learning Algorithm: Returns up to 39.26% in 1 Month
• Home Builders Stocks Based on Machine Learning: Returns up to 42.26% in 3 Months
• SOXX Stocks Based on Deep Learning: Returns up to 681.89% in 1 Year

• Taiwan Semiconductor (TSM) is up 6.83% since July 27th, 2026, as high-performance computing now accounts for 66% of quarterly revenue and full-year 2026 growth is tracking slightly above 40%, with our discounted cash flow model putting fair value at $471.46 per ADR on the strength of its lead across the 2nm through 5nm nodes.

• PayPal Holdings Inc (PYPL) is up 38.94% since June 29th, 2026, as PayPal resets its growth strategy around branded-checkout stabilization and Venmo monetization, with our algorithm showing a high signal on the one-year outlook even as the stock trades well below its DCF fair value.

• Adobe Inc (ADBE) is up 29.41% since June 12th, 2026, on Adobe's recurring subscription model and successful generative AI integration into its creative workflows, which support a case for undervaluation even after this run.

Weekly Winning Forecasts

The AI Portfolio Rebalances This Wednesday: +45.37% Since Inception

Period 18 closes on August 19, when the algorithm selects the next ten positions from scratch. It has been a strong cycle, and Microsoft, one of the names the algorithm held, contributed a 27% gain over the stretch.

• The current period, July 22 to August 19, is up 5.80% against 4.09% for the S&P 500.
• Since inception the portfolio is up 45.37% against 37.65% for the S&P 500, a 7.72 point premium.
• Eighteen rebalance periods completed or in progress since April 2, 2025, tracked with no exclusions.
• Ten stocks and ETFs per cycle, equal weighted, selected entirely by the algorithm and held for roughly 28 days.
• A live and documented track record, not a backtest.

Between Greed and Fear: How AI Strikes the Balance for Good Generalization

Every forecasting model faces the same tension a human investor does: fit the data too closely and you have memorized noise, hold back too much and you have learned nothing. This piece walks through the specific techniques that keep our algorithm on the right side of that line.

• Greed in a model is overfitting, capturing every nuance of the training data. Fear is underfitting, refusing to commit to a pattern at all.
• Principal component analysis isolates the high-variance signal and strips the noise, so the model is not distracted by irrelevant features.
• L1 and L2 regularization constrain complexity, one by forcing weak coefficients to zero, the other by shrinking their magnitude.
• Genetic feature evolution shuffles and recombines feature subsets, tests them against added noise, and keeps only what survives.
• Ensemble methods let many independent models vote, weighted by track record, so no single failure carries the forecast.

How Hedge Funds and Family Offices Use I Know First, and the Performance Behind It

Institutional clients do not use the daily forecasts the way a retail subscriber does, and this piece sets out exactly how they do use them. It also puts the full performance record on the table rather than a selected window of it.

• Funds use the daily signal and predictability readings across 13,500 assets for idea generation, screening and systematic strategy construction.
• The Combined Long/Short Strategy returned 756% since January 29, 2020, against 129.1% for the S&P 500, a compound annual growth rate of 39.7%.
• In 2022 the strategy gained 15.36% while the S&P 500 fell 19.95%, and in 2020 it drew down 9.71% against 33.92% for the index.
• A Sharpe ratio of 1.72 and a Sortino ratio of 2.68 over the full period.
• A J.P. Morgan 2025 survey put active AI use among hedge funds at 46%, up from 18% a year earlier.

Stock Market Forecast: The Top 10 Strategy

The Top 10 Strategy reads a market regime signal continuously and decides whether conditions favor offense or defense before it decides what to hold. The allocation flips with the regime rather than with the calendar.

• In bullish conditions it holds 75% in the top five S&P 500 stocks at 15% each, with the remaining 25% in SPY or RSP by signal strength.
• In bearish conditions it shifts to 25% in Russell 1000 names with the weakest signals at 5% each, and 75% into RSP or SPY defensively.
• Total return of 438.29% from July 2020 through March 15, 2026, exceeding the S&P 500 by 46.09 points.
• A Sharpe ratio of 1.71 and a Sortino ratio of 2.54 over that period.
• Roughly five years and eight months of history, long enough to cover more than one regime.

BKNG Stock Forecast: What Is the Current Value of Global Travel?

Booking Holdings runs a capital-light platform that converts revenue into free cash flow at a rate very few businesses of its size manage. Our discounted cash flow work asks what that engine is actually worth once you discount it honestly.

• The model puts intrinsic value at $215.70 a share, implying 21.5% upside from the July 24, 2026 close of $177.46.
• FY2025 delivered $26.917 billion in revenue and $9.087 billion in free cash flow.
• The model assumes 13.0% constant revenue growth across FY2026E through FY2030E.
• A discount rate of 11.12%, built from the risk-free rate plus beta times the market risk premium.
• Roughly 76.5% of enterprise value sits in terminal value, so small changes to the perpetual growth assumption move the answer a lot.

More:

• AI-Recommended Stocks for August 2026: the full webinar replay
• July 2026 Performance: I Know First Outpaces the S&P 500 as an AI Stock Selloff Splits Hedge Fund Returns
• Introducing the I Know First Leveraged ETF Package

CEO Weekly Letter

Dear Fellow Investors,

The S&P 500 rose 0.36% this week, closing at 7,785.76. The Nasdaq added 0.14% and the Dow actually fell 0.56%, while the Russell 2000 climbed 1.12%. The VIX drifted down to 14.25, its lowest close of the past month.

Read the headline number and you would conclude nothing happened. That reading would be wrong. A week where the small-cap index moves three times as far as the S&P 500 and the Dow moves the other way entirely is not a quiet week. It is a week where the index cancels itself out and the movement all sits underneath, in the individual names. That is precisely the market where a benchmark tells you the least and asset-level forecasting matters the most.

This is a good moment, then, to show you something new we built for exactly that problem.

The Leveraged ETF Package, Live and On the Record

We introduced the Leveraged ETF package two weeks ago. It maps leveraged and inverse instruments onto securities the daily forecasts already cover, 91 stocks, 8 indexes and sectors and 3 commodities across 11 issuers, from 3x long through 3x short. It does not change the forecast. It changes how precisely you can express one.

Its first published forecast is now closed and public, and I want you to see it rather than take my description of it. Over the three sessions from August 13 to August 16, the deep learning forecast covering 20 leveraged ETFs, 10 long and 10 short, returned up to 7.85% on RCAT, with NBIS at 7.13% and SMCI at 5.93%. The package averaged 2.51% against 0.48% for the S&P 500 over the same three days, a 2.03 point premium. Nine of the ten names moved in the direction the algorithm predicted.

Look again at that S&P 500 figure. The index moved 0.48% across those three sessions. The package moved 2.51%. In a flat tape, direction on the right individual names is the entire return, and the leveraged instruments are simply the tool for expressing that direction with precision. I will say the obvious thing plainly, because it is important: these instruments reset daily, their gearing compounds against you in a choppy market, and they are not suitable for every investor or every signal. The package notes say so, and so do I.

When the Signal and the Price Disagree: Netflix

The most useful thing our algorithm does is not confirm what the chart already shows. It is telling you something the chart contradicts. Netflix over the past three months is the clearest example I can give you.


Netflix: the algorithmic signal against the share price. The price falls from mid-May into July while the signal stays positive throughout, then turns higher in early July, ahead of the price.

Netflix fell from roughly $87 in mid-May to $65.08 on July 17, about 25% lower. The signal never followed it down. It stayed positive the whole way, and in early July it began climbing while the price was still falling, hitting one of its strongest readings of the period on the very session Netflix printed its low.

Netflix closed Friday at $78.16, up 20.1% from that low. One name over one stretch is an illustration rather than a track record, but the mechanism is the point: a price chart only tells you where a stock has been.

Where the Algorithm Led This Week

ARK ETF Stocks Forecast Portfolio – 3 Days: Up to 14.49% Return
A self-learning read on the disruptive-innovation universe, up to 14.49% across three sessions.

Top Technology Stocks – 7 Days: Up to 27.91% Return
Genetic algorithms on the technology complex, up to 27.91% over the week.

Quantitative Trading – 14 Days: Up to 32.58% Return
Up to 32.58% over the two-week horizon, an improvement on the same package a week ago.

Tech Stocks – 1 Month: Up to 39.26% Return
Technology again, up to 39.26% over the month, the third technology package in the top six.

Home Builders Stocks – 3 Months: Up to 42.26% Return
Machine learning on home builders, up to 42.26% over the quarter, and worth noting precisely because it is not a technology name.

SOXX Stocks – 1 Year: Up to 681.89% Return
Our deep learning read on the semiconductor complex, up to 681.89% over a 1-year horizon, the single strongest figure across all six time horizons this week.

Top Signals This Cycle

I Know First
Most Popular Package Of The Week

Weekly Apple Stock Update