Robo-advisors: active vs passive investment models

Compounded across a portfolio, it becomes the business model.

Passive platforms dominate automated wealth management because they have aligned three forces that active managers rarely control at the same time: low customer acquisition cost, cheap portfolio construction, and a product that does not need to predict the market. In markets such as Germany, passive robo-advisors are estimated to account for roughly 85% to 90% of the category.

Active robo-advisors are making a different promise. They use tactical asset allocation, proprietary research, direct stock selection, or risk-management signals to move beyond a fixed basket of index ETFs. The proposition is attractive: if the algorithm can generate alpha or reduce drawdowns, a higher fee may be justified. The problem is that the fee is certain, while the alpha is not.

To compare robo-advisor models properly, investors need to look past the interface and examine the underlying economics: what the platform owns, how often it trades, where yield is generated, who carries liquidity risk, and whether the strategy has any realistic chance of covering its fee load.

Why passive indexing owns the category

Passive robo-advisors are not passive in the operational sense. They automate risk profiling, account opening, portfolio construction, tax-aware actions where available, cash management, and rebalancing. What they generally do not automate is a continuous attempt to outguess the market.

The standard passive model starts with a strategic asset allocation. A client may receive a portfolio containing global equity ETFs, government bonds, corporate bonds, and sometimes cash or inflation-sensitive assets. The precise weights depend on the investor’s risk tolerance, investment horizon, and loss capacity. Once those weights are set, the platform periodically rebalances the account back toward its target.

If equities rise sharply and become too large a share of the portfolio, the system sells part of the equity exposure and adds to bonds or cash. If equities fall, it does the reverse. The process is mechanical, transparent, and relatively cheap to operate.

That simplicity is not a weakness. It is the reason the model scales.

A passive robo-advisor does not need a large research team covering individual companies. It does not need to explain every tactical shift to clients. It does not need to maintain a complex signal stack that can fail during a regime change. Its core portfolio can often be implemented with liquid, low-cost ETFs, while the software handles the administrative layer around them.

The commercial advantages are obvious:

• Lower portfolio costs. Broad index ETFs generally have low underlying expense ratios, leaving more of the client’s return exposed to the market.
• Lower trading intensity. Strategic rebalancing produces fewer transactions than a model that constantly rotates between sectors, countries, or individual securities.
• Simple customer communication. The platform can explain what the client owns and why the allocation changes.
• Scalable operations. The same allocation framework can serve thousands of clients with limited incremental investment research.
• Predictable margin structure. If the advisory fee is 0.50% and the operating cost per account is low, the platform has a clearer path to contribution margin.

This is why passive robo-advisors remain the default for most automated investing products. The customer acquisition problem is difficult enough without adding an expensive promise of outperformance.

Passive investing wins the distribution battle because it sells a repeatable process, not a forecast.

The category’s historical development follows the same logic. Vanguard launched the first retail index fund in 1976, establishing the basic product architecture long before the word “robo-advisor” entered financial services. Betterment, founded in 2008, translated that architecture into a digital interface, and Wealthfront followed in 2011. The innovation was not inventing diversified portfolios. It was reducing the friction around them.

That distinction matters for investors and operators. A passive robo-advisor is primarily a distribution and automation business wrapped around low-cost market exposure. Its competitive advantage comes from onboarding, user experience, trust, retention, cash flow, and the ability to attach additional products. The portfolio itself is often the least differentiated component.

What active robo-advisors are actually selling

An active robo-advisor attempts to add a second layer of value: judgment.

That judgment may be encoded in a quantitative allocation model, a machine-learning system, proprietary equity research, or a human investment committee supervising algorithmic recommendations. The platform may alter its exposure to equities and bonds as market conditions change, select ETFs based on expected returns, or own individual stocks and bonds directly.

The crucial difference is not simply that the portfolio changes more often. It is that the platform claims to make allocation decisions that go beyond a fixed market-cap-weighted benchmark.

There are several versions of the active model.

Tactical asset allocation

The platform shifts portfolio weights according to signals such as valuation, momentum, volatility, interest rates, or macroeconomic conditions. A model might reduce equity exposure after a volatility spike or increase exposure to a particular region when valuations appear attractive.

This can help in a narrow set of market environments. It can also produce whipsaw: selling after a decline, buying back after a recovery, and turning a theoretical risk-control mechanism into a source of realized underperformance.

Active ETF selection

Some platforms remain ETF-based but select funds using proprietary research. Zacks Advantage, launched in 2016, is one example of this structure. It uses Zacks research to select ETFs and has been reported to charge approximately 0.35% to 0.70% annually, with a minimum deposit of $25,000.

This model has a more defensible cost profile than an active platform that owns dozens of individual securities, because the underlying vehicles remain diversified and trading can remain relatively controlled. But the investor still needs to ask whether the selection process adds enough value after both the advisory fee and ETF expenses.

Direct stock and bond ownership

Some active robo-advisors bypass ETFs entirely. Solidvest, operated by DJE Kapital, invests directly in individual stocks and bonds and requires a minimum investment of €10,000.

Direct ownership can offer greater control over tax lots, position sizing, and portfolio construction. It may also reduce the layering of ETF fees. But it shifts more responsibility onto the investment process. Security selection, execution quality, liquidity management, corporate actions, and risk concentration all become more important.

Hedge-fund-style portfolios

Titan represents a mobile-first wealth platform using hedge-fund-style portfolios. It manages more than $1.1 billion in assets under management and charges a 0.40% advisory fee, according to the supplied market data.

A platform can call its strategy active without charging a traditional active-management fee. That is one reason the fee range should not be treated as a rigid classification system. The meaningful question is what the investor receives for the fee: broader access, differentiated exposures, downside management, tax benefits, or simply more trading activity.

Active vs passive robo-advisor models

The comparison becomes clearer when the models are placed side by side. Neither is universally superior; they are built around different assumptions about markets, fees, and investor behavior.

The distinction is also behavioral. Passive investing asks the client to accept market volatility in exchange for low implementation costs. Active investing asks the client to trust a process that may look different from the benchmark for extended periods.

That second requirement is more demanding than it sounds. If an active robo-advisor underperforms for six months, the client may tolerate it. If it lags for three years while charging more than a passive alternative, retention becomes a problem. Customer churn then turns an investment thesis into a unit-economics problem.

If the strategy generates durable excess returns, the higher fee can be rational. If it merely creates a more complicated portfolio with similar market exposure, the fee is margin compression disguised as sophistication.

The fee problem is larger than the headline percentage

A robo-advisor’s advisory fee is only the first line item. The actual cost of ownership can include ETF expense ratios, brokerage and execution costs, bid-ask spreads, custody charges, cash drag, foreign-exchange costs, and taxes generated by turnover.

For passive platforms, the cost stack is relatively easy to estimate. A 0.50% advisory fee combined with low-cost ETFs may leave the investor with a transparent and manageable drag on returns. Rebalancing still creates transactions, but the process is usually limited and predictable.

Active platforms face a more difficult equation:

Net return = gross investment return − advisory fee − fund expenses − trading costs − taxes − cash drag

The gross return is uncertain. Most of the deductions are not.

Suppose an active robo-advisor charges 1.00% and uses funds that add another 0.40% in annual expenses. Before considering turnover, spreads, or tax consequences, the strategy must beat a comparable passive portfolio by 1.40 percentage points simply to reach parity. If the active portfolio trades more frequently or holds less liquid securities, the hurdle rises.

That is why the fee difference cannot be discussed as a cosmetic feature. It changes the required investment outcome.

Long-term SPIVA scorecards indicate that approximately 85% to 90% of active fund managers fail to outperform their passive benchmarks after fees over long periods. Robo-advisors are not identical to traditional mutual funds, and their technology may improve execution or tax management. Still, the evidence establishes a demanding baseline: active decisions do not become economically valuable merely because software makes them faster.

The platform has to demonstrate that its process addresses a specific inefficiency or investor constraint. “Our algorithm is intelligent” is not an investment thesis. A credible thesis explains where the return comes from, why competitors cannot arbitrage it away, how much capacity the strategy has, and how the platform behaves when the model is wrong.

A practical cost comparison

The table is intentionally simple. Actual costs vary by provider and account size. But the economic direction does not: active management begins each year with a larger deficit to overcome.

Direct stocks versus ETF-based portfolios

The argument for direct stock ownership is not automatically an argument for active management. It is an argument for a different portfolio-construction method.

ETFs provide immediate diversification and efficient exposure to broad markets. They also compress the operational burden. One trade can provide exposure to hundreds or thousands of securities, and the fund handles internal rebalancing. For a mass-market robo-advisor, that is difficult to beat.

Direct ownership offers more granular control. A platform can avoid certain sectors, manage individual tax lots, express stronger convictions, or combine equity and bond positions in a way that ETFs do not permit. It can also make the portfolio more understandable to investors who want to know exactly which companies they own.

But direct ownership introduces new failure points:

• A portfolio can become concentrated without the client noticing the factor exposure.
• Trading costs may rise when the platform adjusts many individual positions.
• Less liquid securities can create execution slippage.
• Corporate actions and settlement processes add operational complexity.
• The platform’s research quality becomes directly relevant to each holding.
• A model error affects specific securities rather than merely changing ETF weights.

For investors, the right question is not “Are direct stocks better than ETFs?” It is “What control does direct ownership provide, and does that control justify the additional cost and risk?”

For operators, the question is even harsher. Direct portfolios can support higher fees, but they also require more staff, more controls, and greater investment in execution infrastructure. The platform’s gross revenue per account may improve while its cost-to-serve rises. A product that looks premium at the pricing layer can still be unattractive at the contribution-margin level.

Execution is particularly important when an active platform operates in less liquid markets or handles tokenized and crypto-related assets. In those cases, comparing market-making service levels for exchange listings illustrates the kind of operational detail that sits beneath the headline promise of liquidity. The same principle applies to automated portfolios: the displayed allocation is not the whole product; implementation quality determines how much of the theoretical return reaches the client.

Rebalancing: discipline or market timing?

Passive and active robo-advisors both rebalance, but they mean different things by it.

Passive rebalancing is a maintenance function. The portfolio has a target allocation, and the system restores it when market movement causes drift. The objective is not to predict the next market move. It is to prevent the portfolio from gradually taking more risk than the client originally selected.

Active rebalancing is an investment decision. The platform changes the target itself, or moves materially away from it, because its model expects a different risk-return trade-off. That can involve reducing equities before a projected downturn, increasing duration when rates are expected to fall, or rotating toward a sector with stronger expected earnings.

The distinction matters because tactical decisions create timing risk. A passive system can be wrong about the future only in the broad sense that markets may fall. An active system can be wrong twice: first in its forecast, and again in the timing of the trade.

A tactical model may reduce exposure ahead of a decline, but if the market rallies instead, the opportunity cost is immediate. It may then re-enter at higher prices. A model that seeks to limit downside must also explain how it avoids missing the sharp rebounds that frequently follow periods of stress.

Investors should examine the mechanics rather than the marketing language:

1. What triggers a portfolio change? Is the signal based on valuation, momentum, volatility, macroeconomic data, or a discretionary committee?

2. How often can the model trade? A platform that reserves the right to trade daily has a different cost structure from one that reviews allocations monthly or quarterly.

3. Is there a benchmark? Without a clear benchmark, “risk management” can become an unfalsifiable claim.

4. How is success measured? Absolute return, risk-adjusted return, maximum drawdown, or volatility reduction all produce different conclusions.

5. What happens when signals conflict? The difficult cases reveal more than a smooth backtest.

6. Are backtests net of fees and realistic execution costs? A strategy that works before spreads and taxes may not survive in a client account.

A passive robo-advisor usually has a less exciting answer to these questions. That is part of its appeal. The investor knows the portfolio is designed to stay invested rather than make a sequence of discretionary calls.

The performance reality: beating the market is not the same as changing the portfolio

Active robo-advisors often describe their goal as generating alpha or managing downside risk. Those are legitimate objectives, but they are not interchangeable.

Generating alpha means outperforming a relevant benchmark after all costs and with comparable risk. Managing downside risk may mean reducing volatility or limiting drawdowns, even if the portfolio underperforms during a strong bull market. A platform cannot claim success simply because its returns look smoother over a short period.

The benchmark must match the product. An active portfolio holding cash, bonds, and concentrated stocks should not be compared with an all-equity index when the market is rising, then compared with a conservative balanced fund during a drawdown. The comparison needs to reflect the portfolio’s actual risk budget and mandate.

This is where many product claims become vague. A platform may show attractive historical performance without making clear:

• whether returns are gross or net of fees;
• whether the history is live or backtested;
• how survivorship bias was handled;
• whether the benchmark includes dividends;
• whether the portfolio used leverage or derivatives;
• how frequently the model changed;
• and how the strategy behaved in different market regimes.

The absence of long-term, net-of-fee data comparing active robo-advisors with passive platforms is itself significant. Active robo-advisors are relatively recent products, and the category does not yet offer a universal ten-year evidence base. Investors therefore need to place more weight on process transparency and less on short performance windows.

A higher fee is justified only by a measurable advantage. More activity is not a measurable advantage.

For industry professionals, this is also a warning about product positioning. Active management can attract assets during volatile markets because clients become willing to pay for a perception of protection. But assets under management are not proof of investment value. They may reflect distribution, branding, favorable market timing, or a strong mobile experience.

The $1.1 billion managed by Titan demonstrates that an active or hedge-fund-style wealth platform can achieve meaningful scale. It does not, by itself, establish that the strategy has delivered durable alpha after fees. AUM validates product-market fit and distribution efficiency; it does not settle the performance question.

How to choose a robo-advisor without outsourcing judgment

The right choice depends less on whether the platform calls itself active or passive and more on the investor’s actual objective.

A passive robo-advisor is usually the stronger fit when the investor wants:

• broad diversification across global equities and bonds;
• low and predictable costs;
• a rules-based process;
• minimal maintenance;
• long-term accumulation rather than tactical positioning;
• and a portfolio that is easy to benchmark.

An active robo-advisor may be worth investigating when the investor has a specific requirement that passive exposure does not address. That could include direct ownership for tax-lot management, a clearly defined downside-control mandate, access to a differentiated strategy, or a willingness to accept tracking error in pursuit of a stated objective.

The selection process should begin with the portfolio, not the app. Review the asset mix, the instruments used, the historical turnover, the minimum investment, and the total cost. Then ask whether the strategy solves a problem the investor actually has.

A useful comparison framework includes:

• Total cost of ownership: advisory fee, fund expenses, trading costs, foreign exchange, and account charges.
• Portfolio transparency: whether the platform discloses holdings, allocation changes, and rebalancing rules.
• Benchmark discipline: whether performance is compared with an appropriate passive alternative.
• Liquidity: how quickly positions can be traded without material slippage.
• Tax handling: availability of tax-loss harvesting, direct tax-lot control, and reporting quality.
• Minimum investment: active strategies may require substantially larger balances, as seen with Solidvest and Zacks Advantage.
• Cash policy: whether uninvested balances generate yield for the client or become a revenue source for the platform.
• Operational resilience: custody arrangements, execution controls, and the treatment of corporate actions.
• Behavioral fit: whether the investor can tolerate a strategy that deviates from the benchmark for years.

The last point is underestimated. An active model that is theoretically attractive but abandoned after a period of underperformance is not a successful investment strategy for that client. The realized outcome includes the investor’s decision to stay invested.

The business-model test for robo-advisors

For investors, the primary question is net performance relative to risk. For platform operators, the question is whether the product can generate durable contribution margin while delivering a credible outcome.

Passive robo-advisors have an advantage because their cost structure is naturally compatible with scale. They can keep fees low, automate portfolio servicing, and use simple instruments. Their challenge is differentiation. If every competitor offers similar ETF portfolios at similar prices, customer acquisition cost becomes the main battlefield. Distribution partnerships, payroll integration, banking relationships, and cash-management products may matter more than the allocation engine.

Active robo-advisors have a different advantage: differentiation. A proprietary strategy can support a stronger marketing proposition and potentially higher revenue per account. But it must carry research, execution, compliance, customer education, and retention costs. If the active process fails to produce a defensible outcome, the higher fee accelerates churn.

The model survives if the value created exceeds the cost of creating it.

If active management produces a repeatable source of alpha, controls drawdowns without sacrificing too much upside, or delivers tax and customization benefits unavailable through a standard ETF portfolio, then the premium can be economically rational. If it mainly produces more transactions, more opaque explanations, and more expensive exposure to the same market beta, passive products will continue to win.

This is not a prediction about which technology is more advanced. It is a conclusion about incentives. Markets reward platforms that deliver a clear client outcome at a cost they can sustain. Passive robo-advisors have already proven that they can build a scalable product around low-cost exposure. Active robo-advisors still have to prove that their additional layer of judgment is worth paying for after the full cost stack.

For most long-term investors, passive automation remains the default because it minimizes the number of assumptions required to reach a reasonable outcome. Active robo-advisors deserve consideration when they solve a clearly defined portfolio problem, not simply because their interface suggests intelligence.

The market’s final verdict will be made in net returns, retention, and margins. Passive platforms start with the stronger economics. Active platforms survive only if they can turn complexity into measurable value.