Your Portfolio Page — A Step-by-Step Guide

Once the Strategy Builder has turned your idea into a back-tested portfolio,
you save it — and from then on it lives on its own portfolio page. This is
your workbench: it's where you watch what the strategy's model is doing, record
the trades you decide to make, read AI-generated analytical reports, and chat
with the assistant about your numbers.

This guide walks through that page section by section, with one idea threaded
through all of it:

TyBuff is an analytical platform, not a financial adviser. Everything on
this page is an analytical output — a read on what a model is doing under
your saved settings and your chosen instruments. The platform never
places a trade, and it never tells you to. What you do with what you see is
always your call.


Two kinds of portfolios: "AI Analytical" vs "Tracked"

From the navigation you'll see your portfolios split into two columns:

If your page has a TyBuff – A.I. Helper section, you're on an AI Analytical
portfolio. Good — read on.


How you got here

You built a strategy in the Strategy Builder (see The Strategy Builder — A
Step-by-Step Guide
), pressed run, liked the back-test, and saved it. Saving
freezes the strategy's configuration onto the portfolio: its periodicity, its
capital base, its priority rule, and — for a dynamic basket — its materialized
basket schedule. That saved configuration is exactly what the portfolio page
replays and reports on from now on.


Section 1 — Portfolio Data (what the strategy is)

The first card is a plain-language readout of the saved configuration. None of it
is editable here by accident — it's the record of what you built:

For non-daily strategies the page also explains the evaluation window — a
4-day window whose output is timestamped on its closing day — so you can read the
model's timing without guesswork.

If it's a Dynamic Basket

A portfolio built from a Market Explorer ranking shows an extra purple panel:
the ranking used (e.g. TyPulse™), the Top N, the rebalance cadence and
day
, the sector, and whether the universe is Evolving or Fixed. You
can expand View basket schedule to see the exact from → to → tickers windows
that are frozen and replayed. The page is candid here, and so are we:

Because a dynamic basket is selected by past performance, its simulated returns
lean on a handful of historical winners. That's an illustration of the
strategy's mechanics
, not a forecast — and a dynamic basket's Historical
Simulation and Snapshot take longer to compute, so run them once and let them
finish.


Section 2 — Asset Insight (and the freshness check you should actually do)

This card describes the universe the strategy works over.

User-Defined Asset Universe — check your tickers are up to date

Right at the top you'll see User-Defined Asset Universe, the count of assets
the strategy can act on, and — importantly — a data-freshness banner:

Make this a habit. Expand Show tickers and last data date and scan the
column on the right: a green check means the instrument's price data runs through
the most recent trading day; a warning triangle means it's stale. Why it matters:
every model output, snapshot, and analytical figure on this page is only as
current as the data underneath it. The A.I. Helper itself reminds you that its
calculations need data updated through the most recent completed trading day
("yesterday"). If a ticker is lagging, treat any output that depends on it as
provisional until the data catches up.

For a dynamic basket you'll also see the Active Basket — current period (the
instruments eligible in the most recent rebalance window) and a Diversification
breakdown. Note the honest label: diversification here reflects the selection of
instruments, not capital actually invested.


Section 3 — Recording your own trades

This is where you keep the ledger. The portfolio page does not connect to a
broker and does not move money — you record what you are actually doing, and
the page does the arithmetic and keeps the statistics.

Open Positions

Closed Positions

A running history of your last 30 closed trades, each with its realised
Profit / Loss, editable if you mistyped something.

Quick Tools

Inside the Add/Edit dialogs, Quick Tools gives you a Position Sizer, Price
Levels
, % Change, Risk / Reward, and a Currency converter — small
calculators to help you record a position cleanly. They compute; they don't
recommend.

The point of this section: by logging every move you make, the page becomes a
faithful, statistics-rich mirror of your own activity — which is exactly what
makes the AI reports and the chat assistant useful.


Section 4 — The A.I. Helper (see what the model is doing — then decide)

This is the heart of an AI Analytical portfolio. It has two buttons, and the
distinction between them is the whole point.

Model Portfolio Snapshot — what the model is doing right now

Click View Model Allocation and the platform runs your saved strategy against
today's data and shows you the current model output: which instruments the
model is identifying under your settings and your asset universe at this moment.
Each row is tagged:

There's also a Model per-asset allocation (analytical) figure, computed from
your portfolio's current capital basis (deposited capital ± P/L). You can Pin to
side
to keep the snapshot docked on the left edge while you scroll or add
positions.

Read this for what it is: a snapshot of what the model is doing in this moment,
with your configuration.
It is not a buy list. The "Model only" and
"Portfolio only" tags simply show you where your book and the model's current
output differ — whether to align them, and how, is entirely your decision.

Historical Simulation — the back-test, replayed

Click Show me the simulation and the platform replays the back-test of your
saved strategy
— the same TyBuff model(s) assigned to your instruments, run over
history under your saved timing and exposure rules. For a dynamic basket it replays
the frozen basket schedule, so the historical periods never silently change as
new instruments are added to the platform later.

This answers a different question from the snapshot: "How did this exact strategy
behave over the past?"
It's an illustration of mechanics on historical data —
useful context, not a promise about the future.

The mental model to keep

The A.I. Helper lets you observe what the strategy's model is doing — currently
(Snapshot) and historically (Simulation), always with your own settings and your
own universe
. Then you, and only you, decide whether to mirror any of it in
the positions you record. The platform shows; you choose.


Section 5 — AI-Powered Portfolio Analysis (the written reports)

Below the A.I. Helper you can have the AI write up an analytical report on the
data you've recorded. Two depths:

Both produce a written, copy-able report — generated by AI from the data in the
platform at that moment.
As the report's own banner states, automated systems can
produce inaccuracies or omissions; the output is informational and analytical,
not investment advice, and you remain responsible for verifying it and for your own
decisions. Your monthly AI credit usage is shown beneath the buttons.


Section 6 — Chatting with the A.I.

The floating A.I. button (bottom-right) opens two assistants:

Both are powered by large-language-model technology and generate automated,
analytical responses. They do not provide investment advice, recommendations,
trading signals, or personalized guidance, and no advisory relationship is
created
by using them. Treat their answers as a starting point to verify, not a
verdict.


A quick checklist for getting the most out of the page

  1. Check the freshness banner first. Stale tickers make every downstream output
    provisional — update them (or wait for the feed) before leaning on a Snapshot or
    report.
  2. Keep your ledger honest. Record your real adds, edits and closes so the stats,
    reports, and chat reflect your activity.
  3. Use the Snapshot to compare, not to copy. It tells you where your book and the
    model's current output differ — the decision to align is yours.
  4. Read the Simulation as history, not prophecy. A back-test illustrates mechanics;
    it doesn't guarantee anything forward.
  5. Spend AI credits where they help. Reports for written analysis; chat for quick
    questions about your data.

What this page is — and what it isn't

To be completely clear, and consistent with our
Terms & Conditions and
Full Disclaimer:

In short: the portfolio page lets you see, clearly and in your own configuration, what
a model is doing
— currently and historically — and gives you the tools to record and
analyze your own activity around it. Whether to act on any of it is always, and only, up
to you.