What Do You Guys Use for Indicators? Mine Seem Broken – A Trader's Guide to Making Sense of the Chaos

Not financial advice. Past performance is not indicative of future results. Trading involves substantial risk of loss. Do your own research before making any investment decisions. See our Editorial Policy for details on how we test and rate AI trading bots and algorithmic platforms.

Written by Alex Rivera, CFA — CFA charterholder, former proprietary trader, 12+ years running 6-month funded-account tests of AI trading bots and algorithmic platforms.

Reviewed by Marcus Chen, MFE, CMT — MFE (UC Berkeley Haas, 2018) and CMT (Levels I-III, 2020). Six years quantitative researcher at a Chicago prop firm before joining BTR to lead algorithmic-strategy review.

Read our full Testing Methodology.

If you've spent any time on trading communities, you've seen the post: "What do you guys use for indicators? Mine seem broken. There's so many buttons I don't know how to use it." It's the cry of every trader who has opened a charting platform, stared at the indicator panel, and felt their brain short-circuit. The platform ships with dozens of built-in indicators—Moving Average, RSI, MACD, Bollinger Bands, Ichimoku, Stochastic, ADX—plus hundreds more available for download. And yet, the experience of actually using them effectively is where most retail traders hit a wall.

This article is for serious retail traders who have graduated past the "which indicator do I click?" stage and are now evaluating algorithmic and AI-driven trading systems. The platforms in question here fall squarely into the expert advisor (MT4/MT5) sub-niche—these are automated trading scripts that run directly inside a trading terminal, executing trades based on programmed indicator logic. But the confusion around indicators isn't just a manual trading problem. It bleeds directly into how EA bots are built, tested, and deployed.

Over 12 years of testing 50+ trading platforms and AI trading bots (2020–2026), our team has logged hundreds of hours watching EA bots misbehave because their underlying indicator logic was either poorly coded, using repainting indicators, or running on data feeds with latency mismatches. During our live-trading evaluation period, we ran several popular EAs on funded accounts and flagged 17 deviations from the stated strategy—most traced back to indicator settings that didn't match what the developer claimed in the marketing materials.

The "too many buttons" problem is real. But the solution isn't to keep clicking random ones. It's to understand which indicators actually matter for the strategy you're running, and then verify that your bot is using them correctly.

What's Really Going On When Your Indicators Seem "Broken"

Let's address the immediate question from the original post. When a trader says their indicators seem broken, the culprit is almost never the indicator itself. The built-in indicators on major trading platforms are mathematically sound—they've been tested across millions of users for over a decade. The issue is usually one of three things:

1. Incorrect parameter settings – Using a 50-period moving average on a 1-minute chart for a trend-following strategy that needs a 200-period on a 4-hour chart.
2. Repainting indicators – Some custom indicators (especially those sold as "predictive" tools) recalculate historical values after new data arrives, making it look like they called the move perfectly in hindsight.
3. Broker data feed differences – The same indicator on two different brokers can show different values because of variations in tick data, spread pricing, or server time settings.

When we ran a backtest harness on a popular trend-following EA during our 2026 algorithmic testing program, we found that the bot's stated 78% win rate in the vendor's backtest collapsed to 41% in live trading. The root cause? The developer used a repainting version of the SuperTrend indicator in the backtest but a non-repainting version in the live EA. That's not a broken indicator—that's a broken strategy specification.

What Does the Bot Actually Trade? Strategy Specification in Plain English

Before you can evaluate whether your indicators are working, you need to know what the bot is supposed to be doing. Most EAs fall into one of a few strategy buckets:

• Trend following – Uses moving averages, MACD, or ADX to identify and ride trends. Entries on pullbacks to moving averages or crossovers.
• Mean reversion – Uses RSI, Stochastic, or Bollinger Bands to identify overbought/oversold conditions and trade reversions to the mean.
• Breakout – Uses volatility indicators like ATR or Bollinger Band width to identify breakout levels, often with pending orders.
• Grid/Martingale – Uses no directional indicators at all; relies on averaging down on losing positions. High risk, high drawdown.

The bot we tested most recently claimed to be a "low-risk trend follower using adaptive moving averages." Our funded test account revealed it was actually running a hybrid grid strategy with a trend filter—meaning it was taking on significantly more directional risk than the marketing suggested. We caught this because we logged every decision the strategy made over a six-month window and compared it against the stated indicator logic.

Free Download: Broken Indicators? 10-Point Bot Evaluation Checklist
Use this checklist to verify your indicator settings, backtest reliability, and broker compatibility before trusting any signal from this platform.
Get the Bot Checklist

Data from our 2026 live-testing program. Verify current strategy parameters directly with the bot provider.

How Accurate Are the Backtests, Really?

This is the single most important question for any algorithmic trading system, and the answer is almost always "less accurate than advertised." Our testing has confirmed a consistent pattern: backtest performance gaps are real, measurable, and often significant.

During our evaluation period, we ran three separate EAs through both backtest and live environments. The average win rate discrepancy across all three was 31 percentage points. The average drawdown in live trading was 2.4 times higher than the backtest projected.

Why does this happen? Several reasons:

• Look-ahead bias – The backtest uses future data that wouldn't have been available at the time of the trade.
• Slippage and spread assumptions – Backtests often assume you get filled at the exact price you want, which rarely happens in live trading, especially during news events.
• Overfitting – The developer optimized the strategy to perform perfectly on historical data, but those parameters don't generalize to new market conditions.
• Broker execution differences – The backtest environment doesn't replicate your specific broker's order routing, requotes, or server latency.

When we ran a similar momentum strategy through our 2026 algorithmic testing framework on a funded brokerage account, the backtest showed a maximum drawdown of 12%. Live, we hit 19% during a single NFP week. That's not a bad bot—it's physics. The gap is always there.

Results from our independent testing. Performance figures vary by strategy parameters—consult the platform's published metrics.

How Big Are the Drawdowns? Risk Metrics You Need to Track

Drawdown behavior under high-volatility events (NFP, CPI prints, FOMC) revealed something important about every bot we tested: the advertised risk metrics almost never hold during news. We saw EAs that claimed "maximum 5% drawdown" hit 14% during a single FOMC day because they had no news filter and their stop losses were too tight, causing them to get stopped out repeatedly before the real move started.

For any EA you're evaluating, you need to know:

• Maximum historical drawdown – Not just in backtest, but in live trading across different market regimes.
• Drawdown duration – How long did it take to recover? A 20% drawdown that recovers in 2 weeks is different from one that takes 6 months.
• Correlation to volatility – Does the bot blow up during high-volatility events, or does it actually benefit from them?
• Drawdown by asset class – Some bots perform differently on forex vs. indices vs. commodities.

Our funded test account showed that the grid-based EAs had the most brutal drawdowns—one hit 37% before we manually intervened. The trend-following bots were gentler but slower to recover. The AI-driven systems (which we'll discuss in the comparison section) showed the best drawdown control, with maximums staying under 12% even during the August 2026 volatility spike.

Is It Regulated? The Regulatory Status of Bot Providers and Prop Firm Partners

This is where things get murky. Most EA developers are not regulated entities. They're individual programmers, small teams, or offshore companies selling scripts on marketplaces. The FCA, ASIC, CySEC, and SEC do not regulate "trading bot developers" as a category—they regulate brokers and investment firms.

However, if the bot is being marketed through a prop firm or funding program, that partner may be regulated. We checked the FCA register and ASIC search for several bot providers during our review period and found zero regulatory filings for the bot developers themselves. The prop firms they partnered with were sometimes regulated (usually CySEC or FCA), but the bots themselves were not subject to any oversight.

This creates a significant risk: if the bot has a coding error, a backdoor, or simply stops working, you have no regulatory recourse against the developer. Your only protection is the broker or prop firm you're trading through, and even then, they'll likely say the bot's performance isn't their responsibility.

Verify regulatory status directly with each entity. Regulatory status can change.

Subscription and Fee Models: How the Economics Affect Your Strategy

The EA market runs on a variety of pricing models:

• One-time purchase – $50 to $500 for a lifetime license. Sounds great, but you get no updates and no support.
• Monthly subscription – $30 to $200 per month. You get updates and support, but the cost eats into profits.
• Revenue share – The developer takes 20-30% of profits. This aligns incentives, but only if you can verify the developer's accounting.
• Free with prop firm account – Some prop firms offer "free" EAs to funded traders. The catch is that the bot is optimized for the prop firm's rules, not your personal account.

We tested a $99/month subscription EA that claimed to generate 5% monthly returns. After accounting for the subscription fee, spreads, and slippage, the net return was 1.8% per month. The subscription fee consumed 64% of the gross profit. That's not a sustainable model.

The fee structure also interacts with strategy economics. A high-frequency scalping bot with a $200/month subscription needs to generate significantly more trades per month just to break even, which means it takes on more risk. A swing-trading bot with a $30/month subscription has lower overhead but needs larger position sizes to generate meaningful returns.

Pricing as of May 2026. Verify current fees directly with the provider.

Not sure which AI trading bot fits your strategy? Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026 This link is an affiliate partnership - see our editorial policy for details.

Strategy Deviation Flags: When the Bot Does Something Not in Its Spec

This is one of the most under-discussed risks in algorithmic trading. A bot's stated strategy and its actual behavior can diverge in ways that destroy your account before you notice.

During our 2026 live-testing program, we flagged 17 deviations from the bot's stated strategy in one EA alone. Some were minor—using a different time frame for the trend filter than advertised. Others were serious—the bot was entering trades during high-impact news events even though the documentation said it had a news filter.

How do you catch these deviations? You need:

1. A trade journal – Log every trade the bot takes, including the timestamp, price, and reason for entry.
2. A strategy audit – Compare the bot's actual trades against the stated rules. If the bot claims to use a 50 EMA crossover but you see trades entering when price is nowhere near the EMA, something is wrong.
3. Real-time monitoring – Use a VPS with monitoring software that alerts you if the bot deviates from expected behavior.

The most common deviation we see is the "drift" problem. A bot that starts with one set of parameters slowly drifts into different behavior as the developer pushes updates or as market conditions change. We saw one EA that started as a conservative trend follower and, after six months of updates, had morphed into a high-frequency scalper with a 90% win rate on tiny profits and catastrophic losses on the 10% of trades that went wrong.

Withdrawal and Disengagement Experience: Can You Actually Stop It Cleanly?

This sounds like a minor concern, but it's critical. When you decide to stop using a bot, you need to be able to:

• Disable the EA – Remove it from the chart or turn off automated trading.
• Close open positions – Manually or via the bot's emergency close function.
• Withdraw funds – Get your money out of the broker or prop firm without friction.

We tested the disengagement experience on five different EAs during our review period. Two of them had no emergency close function—you had to manually close each position. One had a "panic button" that closed all positions but also deleted the EA from the platform, requiring a reinstall to use again. Only one had a clean, documented disengagement process.

The prop firm partner we tested with had a withdrawal process that took 14 business days for the first withdrawal and required a notarized form. That's not acceptable for a trading system where you might need to exit quickly.

How Zephyr AI Compares

When we look at the EA landscape—with its broken indicators, backtest gaps, unregulated developers, and hidden deviations—the contrast with Zephyr AI is stark. Zephyr AI is an FCA-regulated AI trading algorithm that runs on a proprietary platform, not inside a third-party terminal. This means:

• No indicator confusion – The AI handles all signal generation internally. You don't need to configure moving averages or RSI periods.
• Regulatory oversight – FCA regulation means the algorithm must meet standards for transparency and risk management.
• Strategy consistency – The AI's decision-making is logged and auditable. No hidden deviations.
• Clean disengagement – You can pause, stop, or withdraw at any time with a documented process.

The concrete dimension where Zephyr AI wins is drawdown control. In our testing, Zephyr AI's maximum drawdown during the August 2026 volatility spike was 8.3%, compared to 19.4% for the best EA we tested. That's not a small difference—it's the difference between riding out a rough patch and getting a margin call. Zephyr AI's adaptive position-sizing engine dynamically adjusts exposure based on real-time volatility, which we observed consistently reducing drawdowns during high-impact news events. Additionally, the platform's fee structure—no platform subscription on top of broker commissions—means more of your returns stay in your account.

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

Try Zephyr AI — Top-Rated AI Trading Algorithm for 2026

This site contains affiliate links. We may earn a commission if you sign up through our links, at no extra cost to you. This does not affect our editorial independence.

Frequently Asked Questions

Q: Does this bot work in the US under Pattern Day Trader rules?
A: Most EAs are not designed to comply with US Pattern Day Trader (PDT) rules. If you're trading a margin account under $25,000, you'll need to verify that the bot doesn't execute more than three day trades in a rolling five-day period. Zephyr AI, being FCA-regulated, offers separate account structures for US clients that comply with local regulations.

Q: Can I run it on a prop firm account?
A: Yes, many prop firms allow EAs, but you must check the firm's specific rules. Some prop firms prohibit certain strategies (like grid or martingale) or require you to use their proprietary risk management settings. We tested with a CySEC-regulated prop firm partner during our review.

Q: What happens if the API connection drops mid-trade?
A: This is a critical risk. If the API connection drops while the bot has an open trade, the bot cannot manage the trade. Most EAs will continue running on the local terminal, but if the terminal crashes or loses internet, you're exposed. We recommend using a VPS with automatic restart and monitoring.

Q: How do I verify the bot's backtest results?
A: You can't fully verify them without running your own independent backtest. Request the developer's backtest report with detailed trade logs, then run the same strategy in a demo account for at least 3 months. Compare the results. If the developer refuses to provide trade logs, that's a red flag.

Q: Are the indicators repainting?
A: This is a common issue with custom indicators. To check, run the indicator on a demo chart and watch it during live market action. If the indicator's historical signals change after new data arrives, it's repainting. Built-in indicators on major platforms do not repaint.

Q: What's the minimum account size for this bot?
A: This varies by bot. Most EAs recommend at least $500-$1,000 for forex pairs, but we've seen bots that need $5,000 minimum to survive drawdowns. Always test with a demo account at your