Achieve 5-8% Monthly Returns with Our Institutional ML Trading Algorithms

R-50 is predominantly deployed on Kubera Markets, where it manages multiple hedge funds and family offices. It also supports OX Securities, Vantage Markets, Fusion Markets, and Monaxa. While R-50 is compatible with any FOREX brokerage that satisfies our criteria, we advise using the listed brokerages for the best results

Yes, we do, and it’s quite simple: let our team set up your software (as opposed to doing this yourself), turn it on, and leave it running uninterrupted. If you aren’t profitable in 30-days, we’ll give you your licensing fee back if you request it.

All of our trading solutions are engineered for near-complete automation, requiring minimal manual oversight. However, we recommend regular performance monitoring through the Ridge Capital Systems back office or your brokerage’s platform.

Additionally, clients should perform a weekly verification checklist to ensure uninterrupted connectivity between the platform and their brokerage. This procedure addresses routine brokerage maintenance that may temporarily disrupt connections. The checklist takes approximately 2–3 minutes to complete, and our team provides comprehensive guidance. This task can be performed remotely via your VPS login from any location with internet access.

R-50 and R-400 boast a robust three-year track record, with independently verified, third-party returns available on Myfxbook.

Unlike traditional investments such as stocks or real estate, where returns rely on the appreciation of asset values over time, our algorithms actively extract value from the market through sophisticated machine learning driven strategies. This proactive approach enables us to generate consistent returns regardless of whether the market is rising or falling.

Our strategies are designed to capitalize on market inefficiencies, ensuring that we can perform well in various market conditions. While external global factors can impact trading environments, our advanced risk management protocols effectively mitigate these influences, maintaining the stability and reliability of our performance.

Reflecting the strength of our approach, our algorithms delivered positive returns every month throughout 2023, without a single negative month. In 2024, aside from one month affected by the elections, our performance has continued to remain consistently favorable. This exceptional track record underscores the resilience and effectiveness of our trading solutions in diverse and dynamic market landscapes.

We have never experienced a breakdown in the system. Our central Ai system rides on a cloud-based server which has redundancy built-in. Of course, success also rides on each client’s own VPS and the client ensuring their VPS and brokerage remain connected to their MT4 or MT5. This is an easy task, which takes only a few minutes and we recommend be checked several times per week. We provide full training on how to do so.

R50 and R400 are completely proprietary systems. This does not, however, mean that every strategy the system may elect to employ is based on a proprietary strategy, indicator or even set of indicators. Both systems ride on an intelligent AI “hub” which constantly scans around 150,000 data points across 30 markets, looking for strategies that work in various market conditions (from scalping and intraday to swing positions) and then ranking those positions based on risk/reward categories like sharpe, sortino, omega and calmar ratios. The system will then utilize this database of potential strategies to direct R50 / R400’s sub-systems in the management of individual trade positions to maintain a balanced portfolio of open positions. So while the system itself is proprietary, each individual strategy employed is constantly shifting - some are proprietary and others are not. They are all, however, based on real-life market data points.

Latency is more of a broker question. All connected R50 and R400 systems will be in identical positions at all times. The system itself operates within an extremely nominal latency of less than .001 seconds. However, once an order is pushed out to individual brokers, latency may vary slightly from account to account.

We utilize FXblue and Myfxbook to audit accounts. These provide a live-link between brokerage account and their platforms, and include all historical trade data. We do not show open trades as this would open us up to the risk of copy-trading from these platforms.

Ray Dalio was one of the first people to pioneer the use of computers to aid in decision making, because he realized that computers process data on a neutral plane. They do not make biased decisions. R50 and R400’s central AI system is completely neutral. It looks only at past historical data and what is transpiring in the markets moment-by-moment. It does not have the ability to be biased in its decision making, or make decisions based on a “hunch” or a “hope.” It only sees and reacts to hard-data facts. This is the power of utilizing computers to make decisions: they are incapable of biases and do not have errors in judgement

Because our system is designed to constantly evolve over time, it will continually adjust itself based on what it’s learned. For example, in November of 2024 R-50 saw its first real loss in its near 3-year history (-5.55% for November, 2024). This affected only R-50 (not R-400) and serves as an example of our commitment to constant improvement. The below recaps part of our November Report sent to all R-50 users:

With the election approaching and the Fed interest rate announcement on the calendar, this movement intensified, erasing 4% of EURUSD within just seven days. This so-called correlation breakdown is a rare phenomenon, usually associated with major market shifts and/or black swan events. During such events, the usual correlations collapse into one focal point, in this case, the dollar, making diversification within FX more challenging.

Despite being constructed from mostly uncorrelated strategies (beta < 0.5), four of our universe of 100+ active subsystems were caught up in this, opening exposure on the same side of the USD (short) across multiple pairs. To reduce exposure and mitigate risk, the affected subsystems were initially hedged and later took controlled losses when a stronger move became evident. These subsystems, despite their long and solid uncorrelated track records, experienced a correlation collapse. This is visible in the chart below, which shows the past two years of live performance for these subsystems:

The affected subsystems have been deactivated and will be replaced once volatility returns to a medium-term average. Looking ahead, exposure risk is now significantly reduced, minimizing the potential for further losses.

This was the first time R50 had hit a stop-loss on multiple subsystems in their 2 to 2.5 years of live combined track records. This marked a decline of 4–6% in net performance from October highs, depending on the product, erasing approximately 1.5 months of profits. The negative performance extended across several of our strategies.

Our primary goal at Ridge Capital Solutions is to provide long-term, resilient alpha strategies with limited downside risk. This event demonstrated how effectively our risk management framework handles extraordinary market phases. Although this represents a small setback in terms of overall performance, it underscores our robust risk management methodology, which focuses on controlling exposure and limiting risk.

In reality, sustainable trading requires acknowledging realized losses. We are collecting extensive data on every market movement and corresponding portfolio reactions to continuously enhance our products through real market feedback. A roadmap of near-future improvements is outlined below.

Our roadmap for the coming weeks includes the following. Note that these improvements will be implemented across all existing systems:

- Dynamic correlation/covariance adjustment across subcomponents (single assets traded on subsystems) at 15-minute intervals.
- Implementation of additional trading logic for market regime detection and active trade filtering using
- Hidden Markov Models (HMM) to predict and identify market regime shifts that may lead to correlation breakdowns.
- Introduction of non-mean-reverting strategy pools to replace mean-reverting strategies during trending market phases.
- Reworking subcomponent risk management to prevent clustering of stop-loss events while maintaining tight risk constraints.
- Potential enhancement and introduction of an automated hedging module.
- Significant expansion of engineered features in our machine learning model training files.

R-50 and R-400 have been live in their current state of programming since February, 2022 and August, 2022, respectively. They were not live during the 2020 COVID crash or 2008 financial crisis. However, the design of these systems takes into consideration data available from these time periods, so the system has “learned” from what transpired across 30-markets during 2008 and 2020, although the system was not live at that time.

The worst that can ever happen in any event is hitting position stop losses. As R50 and R400 are not “algorithms” they employ dozens of strategies simultaneously. This is not, for example, a martingale algo that can bring an account from $1 million to zero in an afternoon. On a wider timeframe, we will execute adjustments to R50 and R400’s subsystems whenever we see a need to do so (see example above, regarding an adjustment made to R50 after November, 2024). These do not impact the system’s overall safety protocols in terms of maximum account risk, as R50/400’s AI central server exists to diversify positions and mitigate risk, but to consistently approve on the risk/reward ratios of each individual position taken. We do not build or market software that works on a faulty strategy like a martingale system.

Each individual trade signal and subsystem is monitored by our risk management and filtering module. Real-time net asset exposure and correlation analysis control the current VaR of the portfolio. Our systems operate under rigorous risk limits and are resilient to adverse market conditions, leveraging high-leverage environments up to 1:500 and also providing solutions for leverages as low as 1:3. Any strategy employed by the systems must have a sharpe ratio of at least 2.5, with a focus on achieving an average of 3 across all open positions.

At the lowest scale, each subsystem has its own risk limits per position, e.g. a fixed SL, which is not the same across all trades. R50 has an average risk per subsystem of 2.8%, this means that if all trades of one of the underlying 22 subsystems have to be closed due to the subsystem breaching its DD limit, the impact of that on R50 as a whole is (on average) 2.8%. Of course, the loss per position is much smaller than that. Still, R50 can suffer larger losses than 2.8% if multiple subsystems hit their risk limits in sequence. We develop our systems to be as low correlation with each other as possible to prevent this, but it can and did happen before. Finally, risk is managed at the system level to never permit a DD of more than 30% on R50 as a whole. These responses are similar for R400.

Our contingency plan includes layers of redundancy with real-time backup servers, automated failover systems, and a recovery protocol to ensure minimal downtime. Further, trade execution is segmented by subsystem and in case of a disruption, trades can and will be managed manually. We do have robust cybersecurity measures in place, including segmented sandbox environments for code, encryption, and automated isolation of compromised systems.

Algorithmic strategies work by executing the exact same strategy over and over irrespective of what the market is doing. These can be profitable, but are severely limited in that they will not evolve over time or get smarter. Our goal in building R-50 and R-400 was to employ the benefits of automated, “algorithmic” trading, but build a system that can intelligently manage a portfolio of such positions rather than focusing on a single strategy and hoping the strategy will continue to work into the future. By “correlate with traditional market indices” we assume you mean how do we perform against, for example, simply buying a market-based indice. As a comparison, the US benchmark for market growth between 1957 and 2023 has been an average of 10.26% annually. Our systems are returning between 50.76% and 70.2% annually since their birth.

We have a dedicated tech team monitoring the operation of the AI system 24/7. We are not providing names as we have a strict NDA in place with our development team, but will list their qualifications here.

Our development and oversight team is headed by two individuals who we will call MP and JF.

MP has a biotechnical and aerospace engineering degree from MIT. His experience includes running large scale AI operations in multiple high-stakes fields including brain recovery treatment and space missions in cooperation with NASA and the US dept. of Defense.

JF has a masters degree in finance with focus on asset pricing from Yonsei University / Sogang University, South-Korea. His experience includes the creation and management of technologies powering multiple trading infrastructures across Europe and Asia, including the founding of Germany´s largest copy trading network.

Aligning with our vision and proven best practices in trading and investing, we minimize introducing human decision making into our systematic methodology, as human decision making has proven quite ineffective. Humans are often clouded by emotions, greed, and distractions. This is why Bridgewater became the largest hedge fund in the world: they began using computers to make decisions rather than relying solely on people who are prone to errors in judgement. Computers make decisions only based on data. Human intervention is only permitted if the systems experience a malfunction. However, we do regular audits of all processes within our systems to identify potential shortcomings and initiate counter-measures if necessary.

We do not provide access to the data driving decision making, strategies, or other proprietary information. However, we are happy to provide limited-time access to live accounts via an investor password for any client who wishes to watch the system live for a period of time.

Both R50 and R400 are designed to do what they can to mitigate any concern over liquidity. R50 is intended to manage accounts of between $50,000 - $400,000 USD. Although the system can technically “handle” far more than this, we recommend R400 for accounts with starting balances over $400,000. In FOREX, liquidity really comes down to each broker and their liquidity provider partners. R-400 in particular is designed to execute more, smaller positions to limit exposure to potential liquidity shortfalls. The system has performed quite well market-wide and we have not yet experienced liquidity issues on any broker, even on accounts of over $10 million USD. As for what markets have insufficient liquidity for our systems (omitting the term “strategy” as R50 and R400 are not based around a single “strategy”), the FOREX market is the largest and most liquid market on earth. We would not be able to necessarily trade like this, for example, in the US Options or Futures markets, which have far more limited liquidity. But in FOREX and using an appropriate broker partner, we have no concerns up to and even beyond $25 million per account for R-400 and within $50K - $400k for R-50.

We have no concerns for R-50 between $50,000 and $400,000 USD. R-400 we would have no concerns  for accounts up to $25 million USD. For accounts above $25 million we might suggest running two accounts on different brokerages.

As a higher-end software provider, we offer limited licenses annually. We have no concerns over crowding risk, as we are not selling cheap software to the “masses.” Even if we onboard 500 people per year for R-50 and R-400, this would lead to no discernable impact on the FOREX market or performance capabilities for each system and connected account.

“Ensuring compliance” is a broad question. Without specific details as to what this means, it’s a difficult question to answer. However, broadly, in the United States we remain compliant by offering our products as a Software as Service product. As we are not a hedge fund or investment company, we are not able to manage US client accounts on their behalf. We solve for this by providing each US client with their own software file, installed on their own server/VPS directly to their own brokerage account. This then connects them with our central system, but is fully under each client’s control. In Europe and abroad we operate through a business-to-business partnership with a licensed money-manager, so we can connect client accounts and manage them directly. It is highly unlikely that US law will change anytime soon in any way that will negatively impact our operations. In fact, the coming governmental transition in the US will likely make access to higher-return platforms such as ours easier in the United States, which has remained “behind the times” for some years now.

Yes. We originally developed all of our products to manage our own capital.

We keep no data from clients that would be considered sensitive or “proprietary.” We store only basic contact information and your MT4/5 account number, which is not sensitive information as an MT4 or 5 account number cannot be utilized to make trades on your account or carry out deposits or withdrawals. Only you have access to your account.

We do not have “investors” since we are not an investment company. So we will use the term “Clients” instead. The majority of our current clients are family offices, hedge funds, and high net worth individuals. These individuals and organizations are typically highly intelligent in terms of investing (hence their financial status). They keep the majority of their capital in a low risk/reward category that keeps just ahead of inflation. They they keep increasingly smaller funds with higher risk/reward categories of investments. Any investment that returns 50% or more per year would, by definition, be labeled as “high risk.” Most intelligent investors do not allocate more than 10 - 20% of their funds to the high risk category. For example, an individual with $10 million of investible capital may have $1 - $2 million in the higher risk category. The other $8 - $9 million are kept in increasingly lower risk/reward category investments. It is up to each client to act intelligently with their own funds, and diversify accordingly.

Not to our knowledge. There are few if any true AI systems available to retail end users or even family offices or hedge funds. R-50 and R-400 are the first such publicly available systems we know of, and we are proud to offer them.

Ridge Capital Systems in the United States is a new organization. We have operated for the past 2.5+ years in Europe and abroad only, as we were able to manage client accounts directly there, compliantly. It is only recently we came up with a direct-to-end user software solution for R-50 and R-400. Abroad for these systems we have enjoyed a 100% client retention rate. Not that clients do not diversify over time, but we have not lost a client. In the United States we have only just launched our direct-to-end-user solution and are onboarding a select number of clients in 2025. We anticipate 100% client retention here as well.

This is unrelated to Ridge as a software provider. Each client should choose their broker wisely. Your brokerage holds your funds, not Ridge Capital. Only use reputable brokers who have been in business for a long time.

All R50 and R400 users have access to our private telegram channel where we will announce any system downtime, upgrades, and so on. To date, we have never taken the system fully offline. We issue a monthly report for each system, which is distributed via Telegram as a PDF. This covers the past month’s performance and any adjustments being made to our system(s) as a result of market shifts or lessons learned.

Yes. Every client sets their own risk parameters, which our team will assist with

R-50 and R-400 are intended to be used continually and indefinitely. There is no mandatory “lock up” period (we are not a fund) or any preferred hold time. Each client can decide how to manage their account in terms of deposits and withdrawals. Some clients like to withdraw profits monthly. Others like to compound to seek higher returns over time. Others withdraw quarterly or annually. It’s completely up to you and your financial objectives.