Managed accounts and third party access

What the permission actually covers 

Third party access arrives by one of three routes, and they differ far less than they appear to. Credentials are handed over, so that another person signs in as the account holder and is indistinguishable from them. A written authority is lodged with the firm, naming a person the firm will accept instructions from. Or a technical arrangement is set up so that orders placed on one account are repeated on another automatically. In all three cases the firm receives an instruction it will act on, and in all three cases the instruction did not come from the person whose balance settles it.

What that authority reaches is worth listing, because the pitch for such an arrangement is usually phrased around expertise rather than around permissions. An authority to trade covers which instrument is traded, in which direction, at what size, when a position opens and when it closes. Size is the one that carries the most weight, because size is the multiplier in every profit and loss calculation and it is also what determines how much of the account's equity is committed as margin and therefore no longer available to hold anything else open.

Authorities divide into two kinds, and the distinction has real content rather than being a drafting nicety. A non discretionary authority permits a third party to place orders that the account holder has decided on, so the decision stays in one place and the execution moves. A discretionary authority permits the third party to decide as well, without reference back before each order. Almost every retail arrangement described as managed is discretionary, because the entire proposition is that the account holder does not have to decide anything, and that proposition is only deliverable by handing over the deciding.

A trading authority and a withdrawal authority are separate permissions, and whether they have actually been separated is a question about a specific document rather than a general rule. A narrow trading authority names the account, names the activity and names nothing else, so the party holding it can open and close positions and cannot move money out. A power of attorney drafted broadly, or a general authority signed without its scope being read, can confer both at once. The scope is documentary. It is whatever the document names, and it is knowable before it is signed rather than afterwards.

Who carries the position 

A CFD is written between the client and the firm, and the earlier foundations lesson on what a CFD is sets out why that matters here. A third party who places an order is not a party to the resulting contract. They are the source of an instruction, and nothing more: the contract is still the account holder's, the margin is still taken from the account holder's equity, the running result is still recorded against the account holder's balance, and a close out triggered by that equity failing its requirement is still measured on the account holder's money.

The consequence follows from arithmetic already covered in this academy. A result is calculated on the full notional value of a position rather than on the margin posted against it, so an adverse move is measured against the whole contract, a loss can exhaust the margin entirely, and losses are not limited to the amount deposited. A favourable move is measured on exactly the same basis and to exactly the same degree. None of that changes according to who chose the size. It is a property of the contract, not of the person who typed the order.

The last two rows are the ones the rest of this lesson turns on. The party choosing the instrument, the direction and above all the size posts none of the margin and carries none of the debit. That is not an accusation about anybody's intentions. It is a description of where the exposure sits in the arrangement, and it holds identically whether the party holding the authority is a licensed manager, a relative or a stranger met online.

The licence that managing money requires 

Authorisation is activity specific rather than general, as the first lesson in this module sets out: a register entry names the activities a firm may perform and the products it may perform them on. Managing another person's money at the manager's own discretion is one of those named activities, and it is a different one from executing orders. It carries its own capital requirements, its own conduct rules, its own record and reporting obligations and, in most rulebooks, its own approval of the individuals who perform the function. A firm authorised to execute orders is not thereby authorised to manage them, and neither authorisation implies the other.

That turns a judgement about a person into a lookup with a documentary answer. There is an entity, that entity either appears on the regulator's register or does not, and its entry either lists the management of investments on a discretionary basis among its permitted activities or does not. Individuals are a separate question again: an individual acting in a personal capacity is rarely authorised in their own name, and where a person is permitted to perform the function at all it is normally as an approved individual inside an authorised firm rather than as themselves.

Where money is asked to travel is the second documentary check, and it is the more decisive of the two. A licensed firm is required to hold client money apart from its own, so a request to transfer funds to a personal name, an unrelated company or a payment route with no connection to a licensed entity removes both that separation and the complaints route that goes with it, whatever the paperwork around the request says. YAL holds client funds in segregated accounts, as its regulator requires, and what the regulation of an SCA licensed firm covers is set out in full in the accompanying guide.

The pattern built on the arrangement 

In the sequence a new trader typically meets, this arrangement arrives last, after the calls to follow and the automated system that trades by itself covered in the two preceding lessons. The escalation is orderly from the seller's side. Each step asks for more control and admits less scrutiny: a call can be compared against the chart afterwards, a system's settings can at least be inspected, and an account traded at someone else's discretion produces no visible decision to examine at all until the positions already exist.

The pitch is unusually consistent. No learning is required, no time is required, and payment is described as a share of the profits and nothing else. The last element is what makes the offer feel structurally safe rather than merely attractive, and it is presented that way deliberately: the manager never takes custody of the money, so the arrangement reads as categorically different from transferring funds to a stranger. The money does stay in an account in the account holder's name. What has left is the deciding, and the deciding is what commits the money.

Two versions of the fraudulent form appear repeatedly, and they fail in different places. In the first the access is genuine and so are the trades: credentials or an authority are used to trade the account at a size that produces a run of small recorded gains and then a single loss that takes the equity with it, while a share of each recorded gain is drawn along the way. In the second no trading occurs at all. Money is transferred to be managed, a statement is produced showing positions and a balance, and the arrangement fails at the withdrawal in exactly the shape the earlier lesson on how a trading scam is structured describes. The first leaves a real order history at a real firm. The second leaves a document produced by whoever wrote it.

What withdrawing access does and does not undo 

Ending a third party's access and ending the exposure it created are two separate events, and the gap between them surprises people. Removing an authority stops new instructions from being accepted. It does not close anything already open. Positions that exist continue to move with the market, continue to consume margin against the equity in the account, and remain liable to close out on the same terms as before, until each one is closed by an instruction from someone still permitted to give one.

The mechanics of removal follow the mechanics of the grant, which is a further reason the three routes are worth distinguishing. Access granted through shared credentials ends when those credentials stop working. An authority lodged with the firm is a matter between the account holder and the firm, and ends when the firm is instructed to withdraw it, not when the third party is told. A technical link is severed at the firm as well. Whichever route was used, the firm's own record of the account is unaffected by any of it: an order history exists at the firm, timestamped, listing every instruction the account received, and it does not depend on the cooperation of the person who placed them.

Where practitioners disagree 

Discretionary management of other people's money is an old, licensed and thoroughly supervised activity in the institutional world, and opinion divides sharply over whether the retail version of it is the same thing at a smaller size. One position holds that it plainly is, that the licensing category exists precisely to supervise it, and that treating the whole arrangement as suspect confuses a regulated service with the unlicensed imitations that borrow its vocabulary. The opposing position is that the retail population of these offers is dominated by unlicensed operators to a degree that makes the category practically inseparable from its abuse, so that describing it neutrally understates what a reader is most likely to meet.

A narrower disagreement concerns the performance fee itself. It is defended as the fee structure that aligns a manager with the account, since a manager paid only out of gains earns nothing when there are none. It is criticised on the ground set out above, that a share of gains with no share of losses is an asymmetry regardless of intent, and that flat fees are less flattering to describe but do not vary with the size chosen. Both descriptions are accurate about different aspects of the same contract, which is why the argument persists rather than resolving, and neither of them is a statement about what any particular arrangement will do.